fix: sync macOS voice wake trigger words on blur/submit (#1506 ) (thanks @shiv19)

macOS: fix trigger word input disappearing when typing and on add
Fixed issue where trigger words would disappear when typing or when adding new trigger words. The problem was that `swabbleTriggerWords` changes were triggering `VoiceWakeRuntime.refresh()` which sanitized the array by removing empty strings in real-time. Solution: Introduced local `@State` buffer `triggerEntries` with stable UUID identifiers for each trigger word entry. User edits now only affect the local state buffer and are synced back to `AppState` on explicit actions (submit, remove, disappear). This prevents premature sanitization during editing. The local state is loaded on view appear and when the view becomes active, ensuring it stays in sync with `AppState`.
2026-01-23 19:54:01 +00:00 · 2026-01-23 19:54:01 +00:00 · 2026-01-23 19:53:01 +00:00 · 2026-01-23 19:48:22 +00:00 · 2026-01-23 19:43:26 +00:00 · 2026-01-23 19:43:26 +00:00
367 changed files with 11309 additions and 6330 deletions
@@ -29,5 +29,6 @@ jobs:
          CLAWDBOT_INSTALL_CLI_URL: https://clawd.bot/install-cli.sh
          CLAWDBOT_NO_ONBOARD: "1"
          CLAWDBOT_INSTALL_SMOKE_SKIP_CLI: "1"
+          CLAWDBOT_INSTALL_SMOKE_SKIP_NONROOT: ${{ github.event_name == 'pull_request' && '1' || '0' }}
          CLAWDBOT_INSTALL_SMOKE_PREVIOUS: "2026.1.11-4"
        run: pnpm test:install:smoke
@@ -22,6 +22,15 @@
 - README (GitHub): keep absolute docs URLs (`https://docs.clawd.bot/...`) so links work on GitHub.
 - Docs content must be generic: no personal device names/hostnames/paths; use placeholders like `user@gateway-host` and “gateway host”.

+## exe.dev VM ops (general)
+- Access: SSH to the VM directly: `ssh vm-name.exe.xyz` (or use exe.dev web terminal).
+- Updates: `sudo npm i -g clawdbot@latest` (global install needs root on `/usr/lib/node_modules`).
+- Config: use `clawdbot config set ...`; set `gateway.mode=local` if unset.
+- Restart: exe.dev often lacks systemd user bus; stop old gateway and run:
+  `pkill -9 -f clawdbot-gateway || true; nohup clawdbot gateway run --bind loopback --port 18789 --force > /tmp/clawdbot-gateway.log 2>&1 &`
+- Verify: `clawdbot --version`, `clawdbot health`, `ss -ltnp | rg 18789`.
+- SSH flaky: use exe.dev web terminal or Shelley (web agent) instead of CLI SSH.
+
 ## Build, Test, and Development Commands
 - Runtime baseline: Node **22+** (keep Node + Bun paths working).
 - Install deps: `pnpm install`
@@ -51,6 +60,7 @@
 - Framework: Vitest with V8 coverage thresholds (70% lines/branches/functions/statements).
 - Naming: match source names with `*.test.ts`; e2e in `*.e2e.test.ts`.
 - Run `pnpm test` (or `pnpm test:coverage`) before pushing when you touch logic.
+- Do not set test workers above 16; tried already.
 - Live tests (real keys): `CLAWDBOT_LIVE_TEST=1 pnpm test:live` (Clawdbot-only) or `LIVE=1 pnpm test:live` (includes provider live tests). Docker: `pnpm test:docker:live-models`, `pnpm test:docker:live-gateway`. Onboarding Docker E2E: `pnpm test:docker:onboard`.
 - Full kit + what’s covered: `docs/testing.md`.
 - Pure test additions/fixes generally do **not** need a changelog entry unless they alter user-facing behavior or the user asks for one.
@@ -118,6 +128,10 @@
 - **Multi-agent safety:** do **not** switch branches / check out a different branch unless explicitly requested.
 - **Multi-agent safety:** running multiple agents is OK as long as each agent has its own session.
 - **Multi-agent safety:** when you see unrecognized files, keep going; focus on your changes and commit only those.
+- Lint/format churn:
+  - If staged+unstaged diffs are formatting-only, auto-resolve without asking.
+  - If commit/push already requested, auto-stage and include formatting-only follow-ups in the same commit (or a tiny follow-up commit if needed), no extra confirmation.
+  - Only ask when changes are semantic (logic/data/behavior).
 - Lobster seam: use the shared CLI palette in `src/terminal/palette.ts` (no hardcoded colors); apply palette to onboarding/config prompts and other TTY UI output as needed.
 - **Multi-agent safety:** focus reports on your edits; avoid guard-rail disclaimers unless truly blocked; when multiple agents touch the same file, continue if safe; end with a brief “other files present” note only if relevant.
 - Bug investigations: read source code of relevant npm dependencies and all related local code before concluding; aim for high-confidence root cause.
@@ -2,47 +2,61 @@

 Docs: https://docs.clawd.bot

+## 2026.1.23
+
+### Changes
+- CLI: restart the gateway by default after `clawdbot update`; add `--no-restart` to skip it.
+- CLI: add live auth probes to `clawdbot models status` for per-profile verification.
+- Markdown: add per-channel table conversion (bullets for Signal/WhatsApp, code blocks elsewhere). (#1495) Thanks @odysseus0.
+
+### Fixes
+- macOS: sync voice wake trigger words on submit/blur in settings. (#1506) Thanks @shiv19.
+- TUI: forward unknown slash commands (for example, `/context`) to the Gateway.
+- TUI: include Gateway slash commands in autocomplete and `/help`.
+- CLI: skip usage lines in `clawdbot models status` when provider usage is unavailable.
+- CLI: suppress diagnostic session/run noise during auth probes.
+- Linux: include env-configured user bin roots in systemd PATH and align PATH audits. (#1512) Thanks @robbyczgw-cla.
+- TUI: render Gateway slash-command replies as system output (for example, `/context`).
+- Media: preserve PNG alpha when possible; fall back to JPEG when still over size cap. (#1491) Thanks @robbyczgw-cla.
+- Agents: treat plugin-only tool allowlists as opt-ins; keep core tools enabled. (#1467)
+- Exec approvals: persist allowlist entry ids to keep macOS allowlist rows stable. (#1521) Thanks @ngutman.
+
 ## 2026.1.22

 ### Changes
- Highlight: Lobster optional plugin tool for typed workflows + approval gates. https://docs.clawd.bot/tools/lobster
- Lobster: allow workflow file args via `argsJson` in the plugin tool. https://docs.clawd.bot/tools/lobster
- Agents: add identity avatar config support and Control UI avatar rendering. (#1329, #1424) Thanks @dlauer.
- Memory: prevent CLI hangs by deferring vector probes, adding sqlite-vec/embedding timeouts, and showing sync progress early.
+- Highlight: Compaction safeguard now uses adaptive chunking, progressive fallback, and UI status + retries. (#1466) Thanks @dlauer.
+- Providers: add Antigravity usage tracking to status output. (#1490) Thanks @patelhiren.
+- Slack: add chat-type reply threading overrides via `replyToModeByChatType`. (#1442) Thanks @stefangalescu.
 - BlueBubbles: add `asVoice` support for MP3/CAF voice memos in sendAttachment. (#1477, #1482) Thanks @Nicell.
- Docs: add troubleshooting entry for gateway.mode blocking gateway start. https://docs.clawd.bot/gateway/troubleshooting
- Docs: add /model allowlist troubleshooting note. (#1405)
- Docs: add per-message Gmail search example for gog. (#1220) Thanks @mbelinky.
- UI: show per-session assistant identity in the Control UI. (#1420) Thanks @robbyczgw-cla.
 - Onboarding: add hatch choice (TUI/Web/Later), token explainer, background dashboard seed on macOS, and showcase link.
- Onboarding: remove the run setup-token auth option (paste setup-token or reuse CLI creds instead).
- Signal: add typing indicators and DM read receipts via signal-cli.
- MSTeams: add file uploads, adaptive cards, and attachment handling improvements. (#1410) Thanks @Evizero.
- CLI: add `clawdbot update wizard` for interactive channel selection and restart prompts. https://docs.clawd.bot/cli/update
-
-### Breaking
- **BREAKING:** Envelope and system event timestamps now default to host-local time (was UTC) so agents don’t have to constantly convert.

 ### Fixes
 - BlueBubbles: stop typing indicator on idle/no-reply. (#1439) Thanks @Nicell.
 - Message tool: keep path/filePath as-is for send; hydrate buffers only for sendAttachment. (#1444) Thanks @hopyky.
 - Auto-reply: only report a model switch when session state is available. (#1465) Thanks @robbyczgw-cla.
 - Control UI: resolve local avatar URLs with basePath across injection + identity RPC. (#1457) Thanks @dlauer.
+- Agents: sanitize assistant history text to strip tool-call markers. (#1456) Thanks @zerone0x.
+- Discord: clarify Message Content Intent onboarding hint. (#1487) Thanks @kyleok.
+- Gateway: stop the service before uninstalling and fail if it remains loaded.
 - Agents: surface concrete API error details instead of generic AI service errors.
+- Exec: fall back to non-PTY when PTY spawn fails (EBADF). (#1484)
 - Exec approvals: allow per-segment allowlists for chained shell commands on gateway + node hosts. (#1458) Thanks @czekaj.
 - Agents: make OpenAI sessions image-sanitize-only; gate tool-id/repair sanitization by provider.
 - Doctor: honor CLAWDBOT_GATEWAY_TOKEN for auth checks and security audit token reuse. (#1448) Thanks @azade-c.
 - Agents: make tool summaries more readable and only show optional params when set.
 - Agents: honor SOUL.md guidance even when the file is nested or path-qualified. (#1434) Thanks @neooriginal.
+- Matrix (plugin): persist m.direct for resolved DMs and harden room fallback. (#1436, #1486) Thanks @sibbl.
 - CLI: prefer `~` for home paths in output.
 - Mattermost (plugin): enforce pairing/allowlist gating, keep @username targets, and clarify plugin-only docs. (#1428) Thanks @damoahdominic.
 - Agents: centralize transcript sanitization in the runner; keep <final> tags and error turns intact.
 - Auth: skip auth profiles in cooldown during initial selection and rotation. (#1316) Thanks @odrobnik.
 - Agents/TUI: honor user-pinned auth profiles during cooldown and preserve search picker ranking. (#1432) Thanks @tobiasbischoff.
 - Docs: fix gog auth services example to include docs scope. (#1454) Thanks @zerone0x.
+- Slack: reduce WebClient retries to avoid duplicate sends. (#1481)
 - Slack: read thread replies for message reads when threadId is provided (replies-only). (#1450) Thanks @rodrigouroz.
+- Discord: honor accountId across message actions and cron deliveries. (#1492) Thanks @svkozak.
 - macOS: prefer linked channels in gateway summary to avoid false “not linked” status.
- Providers: improve GitHub Copilot integration (enterprise support, base URL, and auth flow alignment).
+- macOS/tests: fix gateway summary lookup after guard unwrap; prevent browser opens during tests. (ECID-1483)

 ## 2026.1.21-2

@@ -53,6 +67,8 @@ Docs: https://docs.clawd.bot
 ## 2026.1.21

 ### Changes
+- Highlight: Lobster optional plugin tool for typed workflows + approval gates. https://docs.clawd.bot/tools/lobster
+- Lobster: allow workflow file args via `argsJson` in the plugin tool. https://docs.clawd.bot/tools/lobster
 - Heartbeat: allow running heartbeats in an explicit session key. (#1256) Thanks @zknicker.
 - CLI: default exec approvals to the local host, add gateway/node targeting flags, and show target details in allowlist output.
 - CLI: exec approvals mutations render tables instead of raw JSON.
@@ -62,13 +78,24 @@ Docs: https://docs.clawd.bot
 - CLI: flatten node service commands under `clawdbot node` and remove `service node` docs.
 - CLI: move gateway service commands under `clawdbot gateway` and add `gateway probe` for reachability.
 - Sessions: add per-channel reset overrides via `session.resetByChannel`. (#1353) Thanks @cash-echo-bot.
+- Agents: add identity avatar config support and Control UI avatar rendering. (#1329, #1424) Thanks @dlauer.
+- UI: show per-session assistant identity in the Control UI. (#1420) Thanks @robbyczgw-cla.
+- CLI: add `clawdbot update wizard` for interactive channel selection and restart prompts. https://docs.clawd.bot/cli/update
+- Signal: add typing indicators and DM read receipts via signal-cli.
+- MSTeams: add file uploads, adaptive cards, and attachment handling improvements. (#1410) Thanks @Evizero.
+- Onboarding: remove the run setup-token auth option (paste setup-token or reuse CLI creds instead).
+- Docs: add troubleshooting entry for gateway.mode blocking gateway start. https://docs.clawd.bot/gateway/troubleshooting
+- Docs: add /model allowlist troubleshooting note. (#1405)
+- Docs: add per-message Gmail search example for gog. (#1220) Thanks @mbelinky.

 ### Breaking
 - **BREAKING:** Control UI now rejects insecure HTTP without device identity by default. Use HTTPS (Tailscale Serve) or set `gateway.controlUi.allowInsecureAuth: true` to allow token-only auth. https://docs.clawd.bot/web/control-ui#insecure-http
+- **BREAKING:** Envelope and system event timestamps now default to host-local time (was UTC) so agents don’t have to constantly convert.

 ### Fixes
 - Nodes/macOS: prompt on allowlist miss for node exec approvals, persist allowlist decisions, and flatten node invoke errors. (#1394) Thanks @ngutman.
 - Gateway: keep auto bind loopback-first and add explicit tailnet binding to avoid Tailscale taking over local UI. (#1380)
+- Memory: prevent CLI hangs by deferring vector probes, adding sqlite-vec/embedding timeouts, and showing sync progress early.
 - Agents: enforce 9-char alphanumeric tool call ids for Mistral providers. (#1372) Thanks @zerone0x.
 - Embedded runner: persist injected history images so attachments aren’t reloaded each turn. (#1374) Thanks @Nicell.
 - Nodes tool: include agent/node/gateway context in tool failure logs to speed approval debugging.
@@ -2,6 +2,54 @@
 <rss xmlns:sparkle="http://www.andymatuschak.org/xml-namespaces/sparkle" version="2.0">
    <channel>
        <title>Clawdbot</title>
+        <item>
+            <title>2026.1.22</title>
+            <pubDate>Fri, 23 Jan 2026 08:58:14 +0000</pubDate>
+            <link>https://raw.githubusercontent.com/clawdbot/clawdbot/main/appcast.xml</link>
+            <sparkle:version>7530</sparkle:version>
+            <sparkle:shortVersionString>2026.1.22</sparkle:shortVersionString>
+            <sparkle:minimumSystemVersion>15.0</sparkle:minimumSystemVersion>
+            <description><![CDATA[<h2>Clawdbot 2026.1.22</h2>
+<h3>Changes</h3>
+<ul>
+<li>Highlight: Compaction safeguard now uses adaptive chunking, progressive fallback, and UI status + retries. (#1466) Thanks @dlauer.</li>
+<li>Providers: add Antigravity usage tracking to status output. (#1490) Thanks @patelhiren.</li>
+<li>Slack: add chat-type reply threading overrides via <code>replyToModeByChatType</code>. (#1442) Thanks @stefangalescu.</li>
+<li>BlueBubbles: add <code>asVoice</code> support for MP3/CAF voice memos in sendAttachment. (#1477, #1482) Thanks @Nicell.</li>
+<li>Onboarding: add hatch choice (TUI/Web/Later), token explainer, background dashboard seed on macOS, and showcase link.</li>
+</ul>
+<h3>Fixes</h3>
+<ul>
+<li>BlueBubbles: stop typing indicator on idle/no-reply. (#1439) Thanks @Nicell.</li>
+<li>Message tool: keep path/filePath as-is for send; hydrate buffers only for sendAttachment. (#1444) Thanks @hopyky.</li>
+<li>Auto-reply: only report a model switch when session state is available. (#1465) Thanks @robbyczgw-cla.</li>
+<li>Control UI: resolve local avatar URLs with basePath across injection + identity RPC. (#1457) Thanks @dlauer.</li>
+<li>Agents: sanitize assistant history text to strip tool-call markers. (#1456) Thanks @zerone0x.</li>
+<li>Discord: clarify Message Content Intent onboarding hint. (#1487) Thanks @kyleok.</li>
+<li>Gateway: stop the service before uninstalling and fail if it remains loaded.</li>
+<li>Agents: surface concrete API error details instead of generic AI service errors.</li>
+<li>Exec: fall back to non-PTY when PTY spawn fails (EBADF). (#1484)</li>
+<li>Exec approvals: allow per-segment allowlists for chained shell commands on gateway + node hosts. (#1458) Thanks @czekaj.</li>
+<li>Agents: make OpenAI sessions image-sanitize-only; gate tool-id/repair sanitization by provider.</li>
+<li>Doctor: honor CLAWDBOT_GATEWAY_TOKEN for auth checks and security audit token reuse. (#1448) Thanks @azade-c.</li>
+<li>Agents: make tool summaries more readable and only show optional params when set.</li>
+<li>Agents: honor SOUL.md guidance even when the file is nested or path-qualified. (#1434) Thanks @neooriginal.</li>
+<li>Matrix (plugin): persist m.direct for resolved DMs and harden room fallback. (#1436, #1486) Thanks @sibbl.</li>
+<li>CLI: prefer <code>~</code> for home paths in output.</li>
+<li>Mattermost (plugin): enforce pairing/allowlist gating, keep @username targets, and clarify plugin-only docs. (#1428) Thanks @damoahdominic.</li>
+<li>Agents: centralize transcript sanitization in the runner; keep <final> tags and error turns intact.</li>
+<li>Auth: skip auth profiles in cooldown during initial selection and rotation. (#1316) Thanks @odrobnik.</li>
+<li>Agents/TUI: honor user-pinned auth profiles during cooldown and preserve search picker ranking. (#1432) Thanks @tobiasbischoff.</li>
+<li>Docs: fix gog auth services example to include docs scope. (#1454) Thanks @zerone0x.</li>
+<li>Slack: reduce WebClient retries to avoid duplicate sends. (#1481)</li>
+<li>Slack: read thread replies for message reads when threadId is provided (replies-only). (#1450) Thanks @rodrigouroz.</li>
+<li>macOS: prefer linked channels in gateway summary to avoid false “not linked” status.</li>
+<li>macOS/tests: fix gateway summary lookup after guard unwrap; prevent browser opens during tests. (ECID-1483)</li>
+</ul>
+<p><a href="https://github.com/clawdbot/clawdbot/blob/main/CHANGELOG.md">View full changelog</a></p>
+]]></description>
+            <enclosure url="https://github.com/clawdbot/clawdbot/releases/download/v2026.1.22/Clawdbot-2026.1.22.zip" length="22302446" type="application/octet-stream" sparkle:edSignature="w/EzfwGBCRRuCg5vz8enIfYujxOZJWRw9PaunQ7gIafKwnBJSTtxcnkvMVwQsnBwB6VN5Tu2MPij7PjDFFX+CA=="/>
+        </item>
        <item>
            <title>2026.1.21</title>
            <pubDate>Thu, 22 Jan 2026 12:22:35 +0000</pubDate>
@@ -266,21 +314,5 @@ Thanks @AlexMikhalev, @CoreyH, @John-Rood, @KrauseFx, @MaudeBot, @Nachx639, @Nic
 ]]></description>
            <enclosure url="https://github.com/clawdbot/clawdbot/releases/download/v2026.1.21/Clawdbot-2026.1.21.zip" length="12208102" type="application/octet-stream" sparkle:edSignature="hU495Eii8O3qmmUnxYFhXyEGv+qan6KL+GpeuBhPIXf+7B5F/gBh5Oz9cHaqaAPoZ4/3Bo6xgvic0HTkbz6gDw=="/>
        </item>
-        <item>
-            <title>2026.1.16-2</title>
-            <pubDate>Sat, 17 Jan 2026 12:46:22 +0000</pubDate>
-            <link>https://raw.githubusercontent.com/clawdbot/clawdbot/main/appcast.xml</link>
-            <sparkle:version>6273</sparkle:version>
-            <sparkle:shortVersionString>2026.1.16-2</sparkle:shortVersionString>
-            <sparkle:minimumSystemVersion>15.0</sparkle:minimumSystemVersion>
-            <description><![CDATA[<h2>Clawdbot 2026.1.16-2</h2>
-<h3>Changes</h3>
-<ul>
-<li>CLI: stamp build commit into dist metadata so banners show the commit in npm installs.</li>
-</ul>
-<p><a href="https://github.com/clawdbot/clawdbot/blob/main/CHANGELOG.md">View full changelog</a></p>
-]]></description>
-            <enclosure url="https://github.com/clawdbot/clawdbot/releases/download/v2026.1.16-2/Clawdbot-2026.1.16-2.zip" length="21399591" type="application/octet-stream" sparkle:edSignature="zelT+KzN32cXsihbFniPF5Heq0hkwFfL3Agrh/AaoKUkr7kJAFarkGSOZRTWZ9y+DvOluzn2wHHjVigRjMzrBA=="/>
-        </item>
    </channel>
 </rss>
@@ -102,7 +102,9 @@ struct DebugSettings: View {
                        }
                    }

-                Text("When enabled, Clawdbot won't install or manage \(gatewayLaunchdLabel). It will only attach to an existing Gateway.")
+                Text(
+                    "When enabled, Clawdbot won't install or manage \(gatewayLaunchdLabel). " +
+                        "It will only attach to an existing Gateway.")
                    .font(.caption)
                    .foregroundStyle(.secondary)

@@ -84,11 +84,52 @@ enum ExecApprovalDecision: String, Codable, Sendable {
    case deny
 }

-struct ExecAllowlistEntry: Codable, Hashable {
+struct ExecAllowlistEntry: Codable, Hashable, Identifiable {
+    var id: UUID
    var pattern: String
    var lastUsedAt: Double?
    var lastUsedCommand: String?
    var lastResolvedPath: String?
+
+    init(
+        id: UUID = UUID(),
+        pattern: String,
+        lastUsedAt: Double? = nil,
+        lastUsedCommand: String? = nil,
+        lastResolvedPath: String? = nil)
+    {
+        self.id = id
+        self.pattern = pattern
+        self.lastUsedAt = lastUsedAt
+        self.lastUsedCommand = lastUsedCommand
+        self.lastResolvedPath = lastResolvedPath
+    }
+
+    private enum CodingKeys: String, CodingKey {
+        case id
+        case pattern
+        case lastUsedAt
+        case lastUsedCommand
+        case lastResolvedPath
+    }
+
+    init(from decoder: Decoder) throws {
+        let container = try decoder.container(keyedBy: CodingKeys.self)
+        self.id = try container.decodeIfPresent(UUID.self, forKey: .id) ?? UUID()
+        self.pattern = try container.decode(String.self, forKey: .pattern)
+        self.lastUsedAt = try container.decodeIfPresent(Double.self, forKey: .lastUsedAt)
+        self.lastUsedCommand = try container.decodeIfPresent(String.self, forKey: .lastUsedCommand)
+        self.lastResolvedPath = try container.decodeIfPresent(String.self, forKey: .lastResolvedPath)
+    }
+
+    func encode(to encoder: Encoder) throws {
+        var container = encoder.container(keyedBy: CodingKeys.self)
+        try container.encode(self.id, forKey: .id)
+        try container.encode(self.pattern, forKey: .pattern)
+        try container.encodeIfPresent(self.lastUsedAt, forKey: .lastUsedAt)
+        try container.encodeIfPresent(self.lastUsedCommand, forKey: .lastUsedCommand)
+        try container.encodeIfPresent(self.lastResolvedPath, forKey: .lastResolvedPath)
+    }
 }

 struct ExecApprovalsDefaults: Codable {
@@ -295,6 +336,7 @@ enum ExecApprovalsStore {
        let allowlist = ((wildcardEntry.allowlist ?? []) + (agentEntry.allowlist ?? []))
            .map { entry in
                ExecAllowlistEntry(
+                    id: entry.id,
                    pattern: entry.pattern.trimmingCharacters(in: .whitespacesAndNewlines),
                    lastUsedAt: entry.lastUsedAt,
                    lastUsedCommand: entry.lastUsedCommand,
@@ -379,6 +421,7 @@ enum ExecApprovalsStore {
            let allowlist = (entry.allowlist ?? []).map { item -> ExecAllowlistEntry in
                guard item.pattern == pattern else { return item }
                return ExecAllowlistEntry(
+                    id: item.id,
                    pattern: item.pattern,
                    lastUsedAt: Date().timeIntervalSince1970 * 1000,
                    lastUsedCommand: command,
@@ -398,6 +441,7 @@ enum ExecApprovalsStore {
            let cleaned = allowlist
                .map { item in
                    ExecAllowlistEntry(
+                        id: item.id,
                        pattern: item.pattern.trimmingCharacters(in: .whitespacesAndNewlines),
                        lastUsedAt: item.lastUsedAt,
                        lastUsedCommand: item.lastUsedCommand,
@@ -248,12 +248,11 @@ final class GatewayProcessManager {
            guard let linkId else {
                return "port \(port), health probe succeeded, \(instanceText)"
            }
-            let linked = linkId.flatMap { snap.channels[$0]?.linked } ?? false
-            let authAge = linkId.flatMap { snap.channels[$0]?.authAgeMs }.flatMap(msToAge) ?? "unknown age"
+            let linked = snap.channels[linkId]?.linked ?? false
+            let authAge = snap.channels[linkId]?.authAgeMs.flatMap(msToAge) ?? "unknown age"
            let label =
-                linkId.flatMap { snap.channelLabels?[$0] } ??
-                linkId?.capitalized ??
-                "channel"
+                snap.channelLabels?[linkId] ??
+                linkId.capitalized
            let linkText = linked ? "linked" : "not linked"
            return "port \(port), \(label) \(linkText), auth \(authAge), \(instanceText)"
        }
@@ -123,12 +123,12 @@ struct SystemRunSettingsView: View {
                        .foregroundStyle(.secondary)
                } else {
                    VStack(alignment: .leading, spacing: 8) {
-                        ForEach(Array(self.model.entries.enumerated()), id: \.offset) { index, _ in
+                        ForEach(self.model.entries, id: \.id) { entry in
                            ExecAllowlistRow(
                                entry: Binding(
-                                    get: { self.model.entries[index] },
-                                    set: { self.model.updateEntry($0, at: index) }),
-                                onRemove: { self.model.removeEntry(at: index) })
+                                    get: { self.model.entry(for: entry.id) ?? entry },
+                                    set: { self.model.updateEntry($0, id: entry.id) }),
+                                onRemove: { self.model.removeEntry(id: entry.id) })
                        }
                    }
                }
@@ -373,20 +373,24 @@ final class ExecApprovalsSettingsModel {
        ExecApprovalsStore.updateAllowlist(agentId: self.selectedAgentId, allowlist: self.entries)
    }

-    func updateEntry(_ entry: ExecAllowlistEntry, at index: Int) {
+    func updateEntry(_ entry: ExecAllowlistEntry, id: UUID) {
        guard !self.isDefaultsScope else { return }
-        guard self.entries.indices.contains(index) else { return }
+        guard let index = self.entries.firstIndex(where: { $0.id == id }) else { return }
        self.entries[index] = entry
        ExecApprovalsStore.updateAllowlist(agentId: self.selectedAgentId, allowlist: self.entries)
    }

-    func removeEntry(at index: Int) {
+    func removeEntry(id: UUID) {
        guard !self.isDefaultsScope else { return }
-        guard self.entries.indices.contains(index) else { return }
+        guard let index = self.entries.firstIndex(where: { $0.id == id }) else { return }
        self.entries.remove(at: index)
        ExecApprovalsStore.updateAllowlist(agentId: self.selectedAgentId, allowlist: self.entries)
    }

+    func entry(for id: UUID) -> ExecAllowlistEntry? {
+        self.entries.first(where: { $0.id == id })
+    }
+
    func refreshSkillBins(force: Bool = false) async {
        guard self.autoAllowSkills else {
            self.skillBins = []
@@ -21,6 +21,8 @@ struct VoiceWakeSettings: View {
    @State private var micObserver = AudioInputDeviceObserver()
    @State private var micRefreshTask: Task<Void, Never>?
    @State private var availableLocales: [Locale] = []
+    @State private var triggerEntries: [TriggerEntry] = []
+    @FocusState private var focusedTriggerEntryID: UUID?
    private let fieldLabelWidth: CGFloat = 140
    private let controlWidth: CGFloat = 240
    private let isPreview = ProcessInfo.processInfo.isPreview
@@ -31,9 +33,9 @@ struct VoiceWakeSettings: View {
        var id: String { self.uid }
    }

-    private struct IndexedWord: Identifiable {
-        let id: Int
-        let value: String
+    private struct TriggerEntry: Identifiable {
+        let id: UUID
+        var value: String
    }

    private var voiceWakeBinding: Binding<Bool> {
@@ -105,6 +107,7 @@ struct VoiceWakeSettings: View {
        .onAppear {
            guard !self.isPreview else { return }
            self.startMicObserver()
+            self.loadTriggerEntries()
        }
        .onChange(of: self.state.voiceWakeMicID) { _, _ in
            guard !self.isPreview else { return }
@@ -122,8 +125,10 @@ struct VoiceWakeSettings: View {
                self.micRefreshTask = nil
                Task { await self.meter.stop() }
                self.micObserver.stop()
+                self.syncTriggerEntriesToState()
            } else {
                self.startMicObserver()
+                self.loadTriggerEntries()
            }
        }
        .onDisappear {
@@ -136,11 +141,21 @@ struct VoiceWakeSettings: View {
            self.micRefreshTask = nil
            self.micObserver.stop()
            Task { await self.meter.stop() }
+            self.syncTriggerEntriesToState()
+        }
+        .onChange(of: self.state.swabbleTriggerWords) { _, _ in
+            guard !self.isPreview else { return }
+            guard self.focusedTriggerEntryID == nil else { return }
+            self.loadTriggerEntries()
        }
    }

-    private var indexedWords: [IndexedWord] {
-        self.state.swabbleTriggerWords.enumerated().map { IndexedWord(id: $0.offset, value: $0.element) }
+    private func loadTriggerEntries() {
+        self.triggerEntries = self.state.swabbleTriggerWords.map { TriggerEntry(id: UUID(), value: $0) }
+    }
+
+    private func syncTriggerEntriesToState() {
+        self.state.swabbleTriggerWords = self.triggerEntries.map(\.value)
    }

    private var triggerTable: some View {
@@ -154,29 +169,47 @@ struct VoiceWakeSettings: View {
                } label: {
                    Label("Add word", systemImage: "plus")
                }
-                .disabled(self.state.swabbleTriggerWords
-                    .contains(where: { $0.trimmingCharacters(in: .whitespacesAndNewlines).isEmpty }))
+                .disabled(self.triggerEntries
+                    .contains(where: { $0.value.trimmingCharacters(in: .whitespacesAndNewlines).isEmpty }))

-                Button("Reset defaults") { self.state.swabbleTriggerWords = defaultVoiceWakeTriggers }
+                Button("Reset defaults") {
+                    self.triggerEntries = defaultVoiceWakeTriggers.map { TriggerEntry(id: UUID(), value: $0) }
+                    self.syncTriggerEntriesToState()
+                }
            }

-            Table(self.indexedWords) {
-                TableColumn("Word") { row in
-                    TextField("Wake word", text: self.binding(for: row.id))
-                        .textFieldStyle(.roundedBorder)
-                }
-                TableColumn("") { row in
-                    Button {
-                        self.removeWord(at: row.id)
-                    } label: {
-                        Image(systemName: "trash")
+            VStack(spacing: 0) {
+                ForEach(self.$triggerEntries) { $entry in
+                    HStack(spacing: 8) {
+                        TextField("Wake word", text: $entry.value)
+                            .textFieldStyle(.roundedBorder)
+                            .focused(self.$focusedTriggerEntryID, equals: entry.id)
+                            .onSubmit {
+                                self.syncTriggerEntriesToState()
+                            }
+
+                        Button {
+                            self.removeWord(id: entry.id)
+                        } label: {
+                            Image(systemName: "trash")
+                        }
+                        .buttonStyle(.borderless)
+                        .help("Remove trigger word")
+                        .frame(width: 24)
+                    }
+                    .padding(8)
+
+                    if entry.id != self.triggerEntries.last?.id {
+                        Divider()
                    }
-                    .buttonStyle(.borderless)
-                    .help("Remove trigger word")
                }
-                .width(36)
            }
-            .frame(minHeight: 180)
+            .onChange(of: self.focusedTriggerEntryID) { oldValue, newValue in
+                guard oldValue != nil, oldValue != newValue else { return }
+                self.syncTriggerEntriesToState()
+            }
+            .frame(maxWidth: .infinity, minHeight: 180, alignment: .topLeading)
+            .background(Color(nsColor: .textBackgroundColor))
            .clipShape(RoundedRectangle(cornerRadius: 6))
            .overlay(
                RoundedRectangle(cornerRadius: 6)
@@ -211,24 +244,12 @@ struct VoiceWakeSettings: View {
    }

    private func addWord() {
-        self.state.swabbleTriggerWords.append("")
+        self.triggerEntries.append(TriggerEntry(id: UUID(), value: ""))
    }

-    private func removeWord(at index: Int) {
-        guard self.state.swabbleTriggerWords.indices.contains(index) else { return }
-        self.state.swabbleTriggerWords.remove(at: index)
-    }
-
-    private func binding(for index: Int) -> Binding<String> {
-        Binding(
-            get: {
-                guard self.state.swabbleTriggerWords.indices.contains(index) else { return "" }
-                return self.state.swabbleTriggerWords[index]
-            },
-            set: { newValue in
-                guard self.state.swabbleTriggerWords.indices.contains(index) else { return }
-                self.state.swabbleTriggerWords[index] = newValue
-            })
+    private func removeWord(id: UUID) {
+        self.triggerEntries.removeAll { $0.id == id }
+        self.syncTriggerEntriesToState()
    }

    private func toggleTest() {
@@ -638,13 +659,14 @@ extension VoiceWakeSettings {
        state.voicePushToTalkEnabled = true
        state.swabbleTriggerWords = ["Claude", "Hey"]

-        let view = VoiceWakeSettings(state: state, isActive: true)
+        var view = VoiceWakeSettings(state: state, isActive: true)
        view.availableMics = [AudioInputDevice(uid: "mic-1", name: "Built-in")]
        view.availableLocales = [Locale(identifier: "en_US")]
        view.meterLevel = 0.42
        view.meterError = "No input"
        view.testState = .detected("ok")
        view.isTesting = true
+        view.triggerEntries = [TriggerEntry(id: UUID(), value: "Claude")]

        _ = view.body
        _ = view.localePicker
@@ -654,8 +676,9 @@ extension VoiceWakeSettings {
        _ = view.chimeSection

        view.addWord()
-        _ = view.binding(for: 0).wrappedValue
-        view.removeWord(at: 0)
+        if let entryId = view.triggerEntries.first?.id {
+            view.removeWord(id: entryId)
+        }
    }
 }
 #endif
@@ -362,23 +362,72 @@ By default, Clawdbot replies in the main channel. Use `channels.slack.replyToMod

 The mode applies to both auto-replies and agent tool calls (`slack sendMessage`).

-### DM-specific threading
-You can configure different threading behavior for DMs vs channels by setting `channels.slack.dm.replyToMode`:
+### Per-chat-type threading
+You can configure different threading behavior per chat type by setting `channels.slack.replyToModeByChatType`:

 ```json5
 {
  channels: {
    slack: {
      replyToMode: "off",        // default for channels
-      dm: {
-        replyToMode: "all"       // DMs always thread
-      }
+      replyToModeByChatType: {
+        direct: "all",           // DMs always thread
+        group: "first"           // group DMs/MPIM thread first reply
+      },
    }
  }
 }
 ```

-When `dm.replyToMode` is set, DMs use that mode; channels use the top-level `replyToMode`. If `dm.replyToMode` is not set, both DMs and channels use the top-level setting.
+Supported chat types:
+- `direct`: 1:1 DMs (Slack `im`)
+- `group`: group DMs / MPIMs (Slack `mpim`)
+- `channel`: standard channels (public/private)
+
+Precedence:
+1) `replyToModeByChatType.<chatType>`
+2) `replyToMode`
+3) Provider default (`off`)
+
+Legacy `channels.slack.dm.replyToMode` is still accepted as a fallback for `direct` when no chat-type override is set.
+
+Examples:
+
+Thread DMs only:
+```json5
+{
+  channels: {
+    slack: {
+      replyToMode: "off",
+      replyToModeByChatType: { direct: "all" }
+    }
+  }
+}
+```
+
+Thread group DMs but keep channels in the root:
+```json5
+{
+  channels: {
+    slack: {
+      replyToMode: "off",
+      replyToModeByChatType: { group: "first" }
+    }
+  }
+}
+```
+
+Make channels thread, keep DMs in the root:
+```json5
+{
+  channels: {
+    slack: {
+      replyToMode: "first",
+      replyToModeByChatType: { direct: "off", group: "off" }
+    }
+  }
+}
+```

 ### Manual threading tags
 For fine-grained control, use these tags in agent responses:
@@ -700,8 +700,15 @@ Options:
 - `--json`
 - `--plain`
 - `--check` (exit 1=expired/missing, 2=expiring)
+- `--probe` (live probe of configured auth profiles)
+- `--probe-provider <name>`
+- `--probe-profile <id>` (repeat or comma-separated)
+- `--probe-timeout <ms>`
+- `--probe-concurrency <n>`
+- `--probe-max-tokens <n>`

 Always includes the auth overview and OAuth expiry status for profiles in the auth store.
+`--probe` runs live requests (may consume tokens and trigger rate limits).

 ### `models set <model>`
 Set `agents.defaults.model.primary`.
@@ -25,12 +25,26 @@ clawdbot models scan
 `clawdbot models status` shows the resolved default/fallbacks plus an auth overview.
 When provider usage snapshots are available, the OAuth/token status section includes
 provider usage headers.
+Add `--probe` to run live auth probes against each configured provider profile.
+Probes are real requests (may consume tokens and trigger rate limits).

 Notes:
 - `models set <model-or-alias>` accepts `provider/model` or an alias.
 - Model refs are parsed by splitting on the **first** `/`. If the model ID includes `/` (OpenRouter-style), include the provider prefix (example: `openrouter/moonshotai/kimi-k2`).
 - If you omit the provider, Clawdbot treats the input as an alias or a model for the **default provider** (only works when there is no `/` in the model ID).

+### `models status`
+Options:
+- `--json`
+- `--plain`
+- `--check` (exit 1=expired/missing, 2=expiring)
+- `--probe` (live probe of configured auth profiles)
+- `--probe-provider <name>` (probe one provider)
+- `--probe-profile <id>` (repeat or comma-separated profile ids)
+- `--probe-timeout <ms>`
+- `--probe-concurrency <n>`
+- `--probe-max-tokens <n>`
+
 ## Aliases + fallbacks

 ```bash
@@ -1,5 +1,5 @@
 ---
-summary: "CLI reference for `clawdbot update` (safe-ish source update + optional gateway restart)"
+summary: "CLI reference for `clawdbot update` (safe-ish source update + gateway auto-restart)"
 read_when:
  - You want to update a source checkout safely
  - You need to understand `--update` shorthand behavior
@@ -20,14 +20,14 @@ clawdbot update wizard
 clawdbot update --channel beta
 clawdbot update --channel dev
 clawdbot update --tag beta
-clawdbot update --restart
+clawdbot update --no-restart
 clawdbot update --json
 clawdbot --update
 ```

 ## Options

- `--restart`: restart the Gateway service after a successful update.
+- `--no-restart`: skip restarting the Gateway service after a successful update.
 - `--channel <stable|beta|dev>`: set the update channel (git + npm; persisted in config).
 - `--tag <dist-tag|version>`: override the npm dist-tag or version for this update only.
 - `--json`: print machine-readable `UpdateRunResult` JSON.
@@ -52,7 +52,8 @@ Options:
 ## `update wizard`

 Interactive flow to pick an update channel and confirm whether to restart the Gateway
-after updating. If you select `dev` without a git checkout, it offers to create one.
+after updating (default is to restart). If you select `dev` without a git checkout, it
+offers to create one.

 ## What it does

@@ -25,6 +25,7 @@ stay consistent across channels.
 1. **Parse Markdown -> IR**
   - IR is plain text plus style spans (bold/italic/strike/code/spoiler) and link spans.
   - Offsets are UTF-16 code units so Signal style ranges align with its API.
+   - Tables are parsed only when a channel opts into table conversion.
 2. **Chunk IR (format-first)**
   - Chunking happens on the IR text before rendering.
   - Inline formatting does not split across chunks; spans are sliced per chunk.
@@ -59,7 +60,30 @@ IR (schematic):

 - Slack, Telegram, and Signal outbound adapters render from the IR.
 - Other channels (WhatsApp, iMessage, MS Teams, Discord) still use plain text or
-  their own formatting rules.
+  their own formatting rules, with Markdown table conversion applied before
+  chunking when enabled.
+
+## Table handling
+
+Markdown tables are not consistently supported across chat clients. Use
+`markdown.tables` to control conversion per channel (and per account).
+
+- `code`: render tables as code blocks (default for most channels).
+- `bullets`: convert each row into bullet points (default for Signal + WhatsApp).
+- `off`: disable table parsing and conversion; raw table text passes through.
+
+Config keys:
+
+```yaml
+channels:
+  discord:
+    markdown:
+      tables: code
+    accounts:
+      work:
+        markdown:
+          tables: off
+```

 ## Chunking rules

@@ -129,7 +129,7 @@ Save to `~/.clawdbot/clawdbot.json` and you can DM the bot from that number.
        enabled: true,
        maxBytes: 20971520,
        models: [
-          { provider: "openai", model: "whisper-1" },
+          { provider: "openai", model: "gpt-4o-mini-transcribe" },
          // Optional CLI fallback (Whisper binary):
          // { type: "cli", command: "whisper", args: ["--model", "base", "{{MediaPath}}"] }
        ],
@@ -1865,7 +1865,7 @@ Note: `applyPatch` is only under `tools.exec`.
 - Each `models[]` entry:
  - Provider entry (`type: "provider"` or omitted):
    - `provider`: API provider id (`openai`, `anthropic`, `google`/`gemini`, `groq`, etc).
-    - `model`: model id override (required for image; defaults to `whisper-1`/`whisper-large-v3-turbo` for audio providers, and `gemini-3-flash-preview` for video).
+    - `model`: model id override (required for image; defaults to `gpt-4o-mini-transcribe`/`whisper-large-v3-turbo` for audio providers, and `gemini-3-flash-preview` for video).
    - `profile` / `preferredProfile`: auth profile selection.
  - CLI entry (`type: "cli"`):
    - `command`: executable to run.
@@ -1890,7 +1890,7 @@ Example:
          rules: [{ action: "allow", match: { chatType: "direct" } }]
        },
        models: [
-          { provider: "openai", model: "whisper-1" },
+          { provider: "openai", model: "gpt-4o-mini-transcribe" },
          { type: "cli", command: "whisper", args: ["--model", "base", "{{MediaPath}}"] }
        ]
      },
@@ -7,7 +7,7 @@ read_when:

 # Updating

-Clawdbot is moving fast (pre “1.0”). Treat updates like shipping infra: update → run checks → restart → verify.
+Clawdbot is moving fast (pre “1.0”). Treat updates like shipping infra: update → run checks → restart (or use `clawdbot update`, which restarts) → verify.

 ## Recommended: re-run the website installer (upgrade in place)

@@ -81,7 +81,7 @@ Notes:
 For **source installs** (git checkout), prefer:

 ```bash
-clawdbot update --restart
+clawdbot update
 ```

 It runs a safe-ish update flow:
@@ -89,6 +89,7 @@ It runs a safe-ish update flow:
 - Switches to the selected channel (tag or branch).
 - Fetches + rebases against the configured upstream (dev channel).
 - Installs deps, builds, builds the Control UI, and runs `clawdbot doctor`.
+- Restarts the gateway by default (use `--no-restart` to skip).

 If you installed via **npm/pnpm** (no git metadata), `clawdbot update` will try to update via your package manager. If it can’t detect the install, use “Update (global install)” instead.

@@ -6,7 +6,7 @@ read_when:
 # Audio / Voice Notes — 2026-01-17

 ## What works
- **Media understanding (audio)**: If `tools.media.audio` is enabled (or a shared `tools.media.models` entry supports audio), Clawdbot:
+- **Media understanding (audio)**: If audio understanding is enabled (or auto‑detected), Clawdbot:
  1) Locates the first audio attachment (local path or URL) and downloads it if needed.
  2) Enforces `maxBytes` before sending to each model entry.
  3) Runs the first eligible model entry in order (provider or CLI).
@@ -15,6 +15,21 @@ read_when:
 - **Command parsing**: When transcription succeeds, `CommandBody`/`RawBody` are set to the transcript so slash commands still work.
 - **Verbose logging**: In `--verbose`, we log when transcription runs and when it replaces the body.

+## Auto-detection (default)
+If you **don’t configure models** and `tools.media.audio.enabled` is **not** set to `false`,
+Clawdbot auto-detects in this order and stops at the first working option:
+
+1) **Local CLIs** (if installed)
+   - `sherpa-onnx-offline` (requires `SHERPA_ONNX_MODEL_DIR` with encoder/decoder/joiner/tokens)
+   - `whisper-cli` (from `whisper-cpp`; uses `WHISPER_CPP_MODEL` or the bundled tiny model)
+   - `whisper` (Python CLI; downloads models automatically)
+2) **Gemini CLI** (`gemini`) using `read_many_files`
+3) **Provider keys** (OpenAI → Groq → Deepgram → Google)
+
+To disable auto-detection, set `tools.media.audio.enabled: false`.
+To customize, set `tools.media.audio.models`.
+Note: Binary detection is best-effort across macOS/Linux/Windows; ensure the CLI is on `PATH` (we expand `~`), or set an explicit CLI model with a full command path.
+
 ## Config examples

 ### Provider + CLI fallback (OpenAI + Whisper CLI)
@@ -26,7 +41,7 @@ read_when:
        enabled: true,
        maxBytes: 20971520,
        models: [
-          { provider: "openai", model: "whisper-1" },
+          { provider: "openai", model: "gpt-4o-mini-transcribe" },
          {
            type: "cli",
            command: "whisper",
@@ -54,7 +69,7 @@ read_when:
          ]
        },
        models: [
-          { provider: "openai", model: "whisper-1" }
+          { provider: "openai", model: "gpt-4o-mini-transcribe" }
        ]
      }
    }
@@ -83,6 +98,7 @@ read_when:
 - Audio providers can override `baseUrl`, `headers`, and `providerOptions` via `tools.media.audio`.
 - Default size cap is 20MB (`tools.media.audio.maxBytes`). Oversize audio is skipped for that model and the next entry is tried.
 - Default `maxChars` for audio is **unset** (full transcript). Set `tools.media.audio.maxChars` or per-entry `maxChars` to trim output.
+- OpenAI auto default is `gpt-4o-mini-transcribe`; set `model: "gpt-4o-transcribe"` for higher accuracy.
 - Use `tools.media.audio.attachments` to process multiple voice notes (`mode: "all"` + `maxAttachments`).
 - Transcript is available to templates as `{{Transcript}}`.
 - CLI stdout is capped (5MB); keep CLI output concise.
@@ -6,7 +6,7 @@ read_when:
 ---
 # Media Understanding (Inbound) — 2026-01-17

-Clawdbot can optionally **summarize inbound media** (image/audio/video) before the reply pipeline runs. This is **opt-in** and separate from the base attachment flow—if understanding is off, models still receive the original files/URLs as usual.
+Clawdbot can **summarize inbound media** (image/audio/video) before the reply pipeline runs. It auto‑detects when local tools or provider keys are available, and can be disabled or customized. If understanding is off, models still receive the original files/URLs as usual.

 ## Goals
 - Optional: pre‑digest inbound media into short text for faster routing + better command parsing.
@@ -88,6 +88,11 @@ Each `models[]` entry can be **provider** or **CLI**:
 }
 ```

+CLI templates can also use:
+- `{{MediaDir}}` (directory containing the media file)
+- `{{OutputDir}}` (scratch dir created for this run)
+- `{{OutputBase}}` (scratch file base path, no extension)
+
 ## Defaults and limits
 Recommended defaults:
 - `maxChars`: **500** for image/video (short, command‑friendly)
@@ -104,17 +109,22 @@ Rules:
 - If `<capability>.enabled: true` but no models are configured, Clawdbot tries the
  **active reply model** when its provider supports the capability.

-### Auto-enable audio (when keys exist)
-If `tools.media.audio.enabled` is **not** set to `false` and you have any supported
-audio provider keys configured, Clawdbot will **auto-enable audio transcription**
-even when you haven’t listed models explicitly.
+### Auto-detect media understanding (default)
+If `tools.media.<capability>.enabled` is **not** set to `false` and you haven’t
+configured models, Clawdbot auto-detects in this order and **stops at the first
+working option**:

-Providers checked (in order):
-1) OpenAI
-2) Groq
-3) Deepgram
+1) **Local CLIs** (audio only; if installed)
+   - `sherpa-onnx-offline` (requires `SHERPA_ONNX_MODEL_DIR` with encoder/decoder/joiner/tokens)
+   - `whisper-cli` (`whisper-cpp`; uses `WHISPER_CPP_MODEL` or the bundled tiny model)
+   - `whisper` (Python CLI; downloads models automatically)
+2) **Gemini CLI** (`gemini`) using `read_many_files`
+3) **Provider keys**
+   - Audio: OpenAI → Groq → Deepgram → Google
+   - Image: OpenAI → Anthropic → Google → MiniMax
+   - Video: Google

-To disable this behavior, set:
+To disable auto-detection, set:
 ```json5
 {
  tools: {
@@ -126,6 +136,7 @@ To disable this behavior, set:
  }
 }
 ```
+Note: Binary detection is best-effort across macOS/Linux/Windows; ensure the CLI is on `PATH` (we expand `~`), or set an explicit CLI model with a full command path.

 ## Capabilities (optional)
 If you set `capabilities`, the entry only runs for those media types. For shared
@@ -142,7 +153,7 @@ If you omit `capabilities`, the entry is eligible for the list it appears in.
 | Capability | Provider integration | Notes |
 |------------|----------------------|-------|
 | Image | OpenAI / Anthropic / Google / others via `pi-ai` | Any image-capable model in the registry works. |
-| Audio | OpenAI, Groq, Deepgram | Provider transcription (Whisper/Deepgram). |
+| Audio | OpenAI, Groq, Deepgram, Google | Provider transcription (Whisper/Deepgram/Gemini). |
 | Video | Google (Gemini API) | Provider video understanding. |

 ## Recommended providers
@@ -151,8 +162,8 @@ If you omit `capabilities`, the entry is eligible for the list it appears in.
 - Good defaults: `openai/gpt-5.2`, `anthropic/claude-opus-4-5`, `google/gemini-3-pro-preview`.

 **Audio**
- `openai/whisper-1`, `groq/whisper-large-v3-turbo`, or `deepgram/nova-3`.
- CLI fallback: `whisper` binary.
+- `openai/gpt-4o-mini-transcribe`, `groq/whisper-large-v3-turbo`, or `deepgram/nova-3`.
+- CLI fallback: `whisper-cli` (whisper-cpp) or `whisper`.
 - Deepgram setup: [Deepgram (audio transcription)](/providers/deepgram).

 **Video**
@@ -209,7 +220,7 @@ When `mode: "all"`, outputs are labeled `[Image 1/2]`, `[Audio 2/2]`, etc.
      audio: {
        enabled: true,
        models: [
-          { provider: "openai", model: "whisper-1" },
+          { provider: "openai", model: "gpt-4o-mini-transcribe" },
          {
            type: "cli",
            command: "whisper",
@@ -11,7 +11,7 @@ This app now ships Sparkle auto-updates. Release builds must be Developer ID–s

 ## Prereqs
 - Developer ID Application cert installed (example: `Developer ID Application: <Developer Name> (<TEAMID>)`).
- Sparkle private key path set in the environment as `SPARKLE_PRIVATE_KEY_FILE` (path to your Sparkle ed25519 private key; public key baked into Info.plist).
+- Sparkle private key path set in the environment as `SPARKLE_PRIVATE_KEY_FILE` (path to your Sparkle ed25519 private key; public key baked into Info.plist). If it is missing, check `~/.profile`.
 - Notary credentials (keychain profile or API key) for `xcrun notarytool` if you want Gatekeeper-safe DMG/zip distribution.
  - We use a Keychain profile named `clawdbot-notary`, created from App Store Connect API key env vars in your shell profile:
    - `APP_STORE_CONNECT_API_KEY_P8`, `APP_STORE_CONNECT_KEY_ID`, `APP_STORE_CONNECT_ISSUER_ID`
@@ -82,6 +82,8 @@ Enable optional tools in `agents.list[].tools.allow` (or global `tools.allow`):
 ```

 Other config knobs that affect tool availability:
+- Allowlists that only name plugin tools are treated as plugin opt-ins; core tools remain
+  enabled unless you also include core tools or groups in the allowlist.
 - `tools.profile` / `agents.list[].tools.profile` (base allowlist)
 - `tools.byProvider` / `agents.list[].tools.byProvider` (provider‑specific allow/deny)
 - `tools.sandbox.tools.*` (sandbox tool policy when sandboxed)
@@ -16,9 +16,9 @@ provider in two different ways.

 ### 1) Built-in GitHub Copilot provider (`github-copilot`)

-Use the native device-login flow to obtain a GitHub token and use it directly
-against the Copilot API. This is the **default** and simplest path because it
-does not require VS Code. Enterprise domains are supported.
+Use the native device-login flow to obtain a GitHub token, then exchange it for
+Copilot API tokens when Clawdbot runs. This is the **default** and simplest path
+because it does not require VS Code.

 ### 2) Copilot Proxy plugin (`copilot-proxy`)

@@ -39,8 +39,6 @@ clawdbot models auth login-github-copilot

 You'll be prompted to visit a URL and enter a one-time code. Keep the terminal
 open until it completes.
-If you're on GitHub Enterprise, the login will ask for your enterprise URL or
-domain (for example `company.ghe.com`).

 ### Optional flags

@@ -68,7 +66,5 @@ clawdbot models set github-copilot/gpt-4o
 - Requires an interactive TTY; run it directly in a terminal.
 - Copilot model availability depends on your plan; if a model is rejected, try
  another ID (for example `github-copilot/gpt-4.1`).
- The login stores a GitHub token in the auth profile store and uses it directly
-  for Copilot API calls.
- Base URL: `https://api.githubcopilot.com` (public) or `https://copilot-api.<domain>`
-  for GitHub Enterprise.
+- The login stores a GitHub token in the auth profile store and exchanges it for a
+  Copilot API token when Clawdbot runs.
@@ -13,7 +13,7 @@ Use `pnpm` (Node 22+) from the repo root. Keep the working tree clean before tag
 ## Operator trigger
 When the operator says “release”, immediately do this preflight (no extra questions unless blocked):
 - Read this doc and `docs/platforms/mac/release.md`.
- Load env from `~/.profile` and confirm `SPARKLE_PRIVATE_KEY_FILE` + App Store Connect vars are set.
+- Load env from `~/.profile` and confirm `SPARKLE_PRIVATE_KEY_FILE` + App Store Connect vars are set (SPARKLE_PRIVATE_KEY_FILE should live in `~/.profile`).
 - Use Sparkle keys from `~/Library/CloudStorage/Dropbox/Backup/Sparkle` if needed.

 1) **Version & metadata**
@@ -39,8 +39,9 @@ When the operator says “release”, immediately do this preflight (no extra qu
 - [ ] `pnpm test` (or `pnpm test:coverage` if you need coverage output)
 - [ ] `pnpm run build` (last sanity check after tests)
 - [ ] `pnpm release:check` (verifies npm pack contents)
- [ ] `pnpm test:install:smoke` (Docker install smoke test; required before release)
+- [ ] `CLAWDBOT_INSTALL_SMOKE_SKIP_NONROOT=1 pnpm test:install:smoke` (Docker install smoke test, fast path; required before release)
  - If the immediate previous npm release is known broken, set `CLAWDBOT_INSTALL_SMOKE_PREVIOUS=<last-good-version>` or `CLAWDBOT_INSTALL_SMOKE_SKIP_PREVIOUS=1` for the preinstall step.
+- [ ] (Optional) Full installer smoke (adds non-root + CLI coverage): `pnpm test:install:smoke`
 - [ ] (Optional) Installer E2E (Docker, runs `curl -fsSL https://clawd.bot/install.sh | bash`, onboards, then runs real tool calls):
  - `pnpm test:install:e2e:openai` (requires `OPENAI_API_KEY`)
  - `pnpm test:install:e2e:anthropic` (requires `ANTHROPIC_API_KEY`)
@@ -324,6 +324,7 @@ brew install <formula>
 ```

 If you run Clawdbot via systemd, ensure the service PATH includes `/home/linuxbrew/.linuxbrew/bin` (or your brew prefix) so `brew`-installed tools resolve in non‑login shells.
+Recent builds also prepend common user bin dirs on Linux systemd services (for example `~/.local/bin`, `~/.npm-global/bin`, `~/.local/share/pnpm`, `~/.bun/bin`) and honor `PNPM_HOME`, `NPM_CONFIG_PREFIX`, `BUN_INSTALL`, `VOLTA_HOME`, `ASDF_DATA_DIR`, `NVM_DIR`, and `FNM_DIR` when set.

 ### Can I switch between npm and git installs later?

@@ -54,6 +54,7 @@ Example schema:
      "autoAllowSkills": true,
      "allowlist": [
        {
+          "id": "B0C8C0B3-2C2D-4F8A-9A3C-5A4B3C2D1E0F",
          "pattern": "~/Projects/**/bin/rg",
          "lastUsedAt": 1737150000000,
          "lastUsedCommand": "rg -n TODO",
@@ -96,6 +97,7 @@ Examples:
 - `/opt/homebrew/bin/rg`

 Each allowlist entry tracks:
+- **id** stable UUID used for UI identity (optional)
 - **last used** timestamp
 - **last used command**
 - **last resolved path**
@@ -121,6 +121,10 @@ Lobster is an **optional** plugin tool (not enabled by default). Allow it per ag

 You can also allow it globally with `tools.allow` if every agent should see it.

+Note: allowlists are opt-in for optional plugins. If your allowlist only names
+plugin tools (like `lobster`), Clawdbot keeps core tools enabled. To restrict core
+tools, include the core tools or groups you want in the allowlist too.
+
 ## Example: Email triage

 Without Lobster:
@@ -88,6 +88,8 @@ Session lifecycle:
 - `/settings`
 - `/exit`

+Other Gateway slash commands (for example, `/context`) are forwarded to the Gateway and shown as system output. See [Slash commands](/tools/slash-commands).
+
 ## Local shell commands
 - Prefix a line with `!` to run a local shell command on the TUI host.
 - The TUI prompts once per session to allow local execution; declining keeps `!` disabled for the session.
@@ -1,3 +1,4 @@
+import { MarkdownConfigSchema } from "clawdbot/plugin-sdk";
 import { z } from "zod";

 const allowFromEntry = z.union([z.string(), z.number()]);
@@ -25,6 +26,7 @@ const bluebubblesGroupConfigSchema = z.object({
 const bluebubblesAccountSchema = z.object({
  name: z.string().optional(),
  enabled: z.boolean().optional(),
+  markdown: MarkdownConfigSchema,
  serverUrl: z.string().optional(),
  password: z.string().optional(),
  webhookPath: z.string().optional(),
@@ -99,6 +99,8 @@ function createMockRuntime(): PluginRuntime {
        chunkText: vi.fn() as unknown as PluginRuntime["channel"]["text"]["chunkText"],
        resolveTextChunkLimit: vi.fn(() => 4000) as unknown as PluginRuntime["channel"]["text"]["resolveTextChunkLimit"],
        hasControlCommand: mockHasControlCommand as unknown as PluginRuntime["channel"]["text"]["hasControlCommand"],
+        resolveMarkdownTableMode: vi.fn(() => "code") as unknown as PluginRuntime["channel"]["text"]["resolveMarkdownTableMode"],
+        convertMarkdownTables: vi.fn((text: string) => text) as unknown as PluginRuntime["channel"]["text"]["convertMarkdownTables"],
      },
      reply: {
        dispatchReplyWithBufferedBlockDispatcher: mockDispatchReplyWithBufferedBlockDispatcher as unknown as PluginRuntime["channel"]["reply"]["dispatchReplyWithBufferedBlockDispatcher"],
@@ -220,6 +222,12 @@ function createMockResponse(): ServerResponse & { body: string; statusCode: numb
  return res;
 }

+const flushAsync = async () => {
+  for (let i = 0; i < 2; i += 1) {
+    await new Promise<void>((resolve) => setImmediate(resolve));
+  }
+};
+
 describe("BlueBubbles webhook monitor", () => {
  let unregister: () => void;

@@ -506,7 +514,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(resolveChatGuidForTarget).toHaveBeenCalledWith(
        expect.objectContaining({
@@ -554,7 +562,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(resolveChatGuidForTarget).not.toHaveBeenCalled();
      expect(sendMessageBlueBubbles).toHaveBeenCalledWith(
@@ -601,7 +609,7 @@ describe("BlueBubbles webhook monitor", () => {
      await handleBlueBubblesWebhookRequest(req, res);

      // Wait for async processing
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(res.statusCode).toBe(200);
      expect(mockDispatchReplyWithBufferedBlockDispatcher).toHaveBeenCalled();
@@ -640,7 +648,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(res.statusCode).toBe(200);
      expect(mockDispatchReplyWithBufferedBlockDispatcher).not.toHaveBeenCalled();
@@ -681,7 +689,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(mockUpsertPairingRequest).toHaveBeenCalled();
      expect(mockDispatchReplyWithBufferedBlockDispatcher).not.toHaveBeenCalled();
@@ -724,7 +732,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(mockUpsertPairingRequest).toHaveBeenCalled();
      // Should not send pairing reply since created=false
@@ -765,7 +773,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(mockDispatchReplyWithBufferedBlockDispatcher).toHaveBeenCalled();
    });
@@ -802,7 +810,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(mockDispatchReplyWithBufferedBlockDispatcher).not.toHaveBeenCalled();
    });
@@ -842,7 +850,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(mockDispatchReplyWithBufferedBlockDispatcher).toHaveBeenCalled();
    });
@@ -880,7 +888,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(mockDispatchReplyWithBufferedBlockDispatcher).not.toHaveBeenCalled();
    });
@@ -919,7 +927,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(mockDispatchReplyWithBufferedBlockDispatcher).not.toHaveBeenCalled();
    });
@@ -958,7 +966,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(mockDispatchReplyWithBufferedBlockDispatcher).toHaveBeenCalled();
    });
@@ -999,7 +1007,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(mockDispatchReplyWithBufferedBlockDispatcher).toHaveBeenCalled();
      const callArgs = mockDispatchReplyWithBufferedBlockDispatcher.mock.calls[0][0];
@@ -1040,7 +1048,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(mockDispatchReplyWithBufferedBlockDispatcher).not.toHaveBeenCalled();
    });
@@ -1078,7 +1086,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(mockDispatchReplyWithBufferedBlockDispatcher).toHaveBeenCalled();
    });
@@ -1121,7 +1129,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(mockDispatchReplyWithBufferedBlockDispatcher).toHaveBeenCalled();
      const callArgs = mockDispatchReplyWithBufferedBlockDispatcher.mock.calls[0][0];
@@ -1167,7 +1175,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(mockDispatchReplyWithBufferedBlockDispatcher).toHaveBeenCalled();
      const callArgs = mockDispatchReplyWithBufferedBlockDispatcher.mock.calls[0][0];
@@ -1213,7 +1221,7 @@ describe("BlueBubbles webhook monitor", () => {
      const originalRes = createMockResponse();

      await handleBlueBubblesWebhookRequest(originalReq, originalRes);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      // Only assert the reply message behavior below.
      mockDispatchReplyWithBufferedBlockDispatcher.mockClear();
@@ -1237,7 +1245,7 @@ describe("BlueBubbles webhook monitor", () => {
      const replyRes = createMockResponse();

      await handleBlueBubblesWebhookRequest(replyReq, replyRes);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(mockDispatchReplyWithBufferedBlockDispatcher).toHaveBeenCalled();
      const callArgs = mockDispatchReplyWithBufferedBlockDispatcher.mock.calls[0][0];
@@ -1283,7 +1291,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(mockDispatchReplyWithBufferedBlockDispatcher).toHaveBeenCalled();
      const callArgs = mockDispatchReplyWithBufferedBlockDispatcher.mock.calls[0][0];
@@ -1331,7 +1339,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(sendBlueBubblesReaction).toHaveBeenCalledWith(
        expect.objectContaining({
@@ -1384,7 +1392,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      // Should process even without mention because it's an authorized control command
      expect(mockDispatchReplyWithBufferedBlockDispatcher).toHaveBeenCalled();
@@ -1427,7 +1435,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(mockDispatchReplyWithBufferedBlockDispatcher).not.toHaveBeenCalled();
    });
@@ -1470,7 +1478,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(markBlueBubblesChatRead).toHaveBeenCalled();
    });
@@ -1511,7 +1519,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(markBlueBubblesChatRead).not.toHaveBeenCalled();
    });
@@ -1554,7 +1562,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      // Should call typing start when reply flow triggers it.
      expect(sendBlueBubblesTyping).toHaveBeenCalledWith(
@@ -1604,7 +1612,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(sendBlueBubblesTyping).toHaveBeenCalledWith(
        expect.any(String),
@@ -1649,7 +1657,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(sendBlueBubblesTyping).toHaveBeenCalledWith(
        expect.any(String),
@@ -1697,7 +1705,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      // Outbound message ID uses short ID "2" (inbound msg-1 is "1", outbound msg-123 is "2")
      expect(mockEnqueueSystemEvent).toHaveBeenCalledWith(
@@ -1742,7 +1750,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(mockEnqueueSystemEvent).toHaveBeenCalledWith(
        expect.stringContaining("reaction added"),
@@ -1782,7 +1790,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(mockEnqueueSystemEvent).toHaveBeenCalledWith(
        expect.stringContaining("reaction removed"),
@@ -1822,7 +1830,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(mockEnqueueSystemEvent).not.toHaveBeenCalled();
    });
@@ -1860,7 +1868,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(mockEnqueueSystemEvent).toHaveBeenCalledWith(
        expect.stringContaining("👍"),
@@ -1901,7 +1909,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(mockDispatchReplyWithBufferedBlockDispatcher).toHaveBeenCalled();
      const callArgs = mockDispatchReplyWithBufferedBlockDispatcher.mock.calls[0][0];
@@ -1941,7 +1949,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      // The short ID "1" should resolve back to the full UUID
      expect(resolveBlueBubblesMessageId("1")).toBe("msg-uuid-12345");
@@ -1993,7 +2001,7 @@ describe("BlueBubbles webhook monitor", () => {
      const res = createMockResponse();

      await handleBlueBubblesWebhookRequest(req, res);
-      await new Promise((resolve) => setTimeout(resolve, 50));
+      await flushAsync();

      expect(mockDispatchReplyWithBufferedBlockDispatcher).not.toHaveBeenCalled();
    });
@@ -1662,9 +1662,15 @@ async function processMessage(
              ? [payload.mediaUrl]
              : [];
          if (mediaList.length > 0) {
+            const tableMode = core.channel.text.resolveMarkdownTableMode({
+              cfg: config,
+              channel: "bluebubbles",
+              accountId: account.accountId,
+            });
+            const text = core.channel.text.convertMarkdownTables(payload.text ?? "", tableMode);
            let first = true;
            for (const mediaUrl of mediaList) {
-              const caption = first ? payload.text : undefined;
+              const caption = first ? text : undefined;
              first = false;
              const result = await sendBlueBubblesMedia({
                cfg: config,
@@ -1686,8 +1692,14 @@ async function processMessage(
            account.config.textChunkLimit && account.config.textChunkLimit > 0
              ? account.config.textChunkLimit
              : DEFAULT_TEXT_LIMIT;
-          const chunks = core.channel.text.chunkMarkdownText(payload.text ?? "", textLimit);
-          if (!chunks.length && payload.text) chunks.push(payload.text);
+          const tableMode = core.channel.text.resolveMarkdownTableMode({
+            cfg: config,
+            channel: "bluebubbles",
+            accountId: account.accountId,
+          });
+          const text = core.channel.text.convertMarkdownTables(payload.text ?? "", tableMode);
+          const chunks = core.channel.text.chunkMarkdownText(text, textLimit);
+          if (!chunks.length && text) chunks.push(text);
          if (!chunks.length) return;
          for (const chunk of chunks) {
            const result = await sendMessageBlueBubbles(outboundTarget, chunk, {
@@ -1,3 +1,4 @@
+import { MarkdownConfigSchema } from "clawdbot/plugin-sdk";
 import { z } from "zod";

 const allowFromEntry = z.union([z.string(), z.number()]);
@@ -35,6 +36,7 @@ const matrixRoomSchema = z
 export const MatrixConfigSchema = z.object({
  name: z.string().optional(),
  enabled: z.boolean().optional(),
+  markdown: MarkdownConfigSchema,
  homeserver: z.string().optional(),
  userId: z.string().optional(),
  accessToken: z.string().optional(),
@@ -53,7 +53,7 @@ export async function resolveMatrixAuth(params?: {
    saveMatrixCredentials,
    credentialsMatchConfig,
    touchMatrixCredentials,
-  } = await import("./credentials.js");
+  } = await import("../credentials.js");

  const cached = loadMatrixCredentials(env);
  const cachedCredentials =
@@ -548,6 +548,11 @@ export function createMatrixRoomMessageHandler(params: MatrixMonitorHandlerParam
      }

      let didSendReply = false;
+      const tableMode = core.channel.text.resolveMarkdownTableMode({
+        cfg,
+        channel: "matrix",
+        accountId: route.accountId,
+      });
      const { dispatcher, replyOptions, markDispatchIdle } =
        core.channel.reply.createReplyDispatcherWithTyping({
          responsePrefix: core.channel.reply.resolveEffectiveMessagesConfig(cfg, route.agentId)
@@ -562,6 +567,8 @@ export function createMatrixRoomMessageHandler(params: MatrixMonitorHandlerParam
              textLimit,
              replyToMode,
              threadId: threadTarget,
+              accountId: route.accountId,
+              tableMode,
            });
            didSendReply = true;
          },
@@ -1,6 +1,6 @@
 import type { MatrixClient } from "matrix-bot-sdk";

-import type { ReplyPayload, RuntimeEnv } from "clawdbot/plugin-sdk";
+import type { MarkdownTableMode, ReplyPayload, RuntimeEnv } from "clawdbot/plugin-sdk";
 import { sendMessageMatrix } from "../send.js";
 import { getMatrixRuntime } from "../../runtime.js";

@@ -12,8 +12,17 @@ export async function deliverMatrixReplies(params: {
  textLimit: number;
  replyToMode: "off" | "first" | "all";
  threadId?: string;
+  accountId?: string;
+  tableMode?: MarkdownTableMode;
 }): Promise<void> {
  const core = getMatrixRuntime();
+  const tableMode =
+    params.tableMode ??
+    core.channel.text.resolveMarkdownTableMode({
+      cfg: core.config.loadConfig(),
+      channel: "matrix",
+      accountId: params.accountId,
+    });
  const logVerbose = (message: string) => {
    if (core.logging.shouldLogVerbose()) {
      params.runtime.log?.(message);
@@ -33,6 +42,8 @@ export async function deliverMatrixReplies(params: {
    }
    const replyToIdRaw = reply.replyToId?.trim();
    const replyToId = params.threadId || params.replyToMode === "off" ? undefined : replyToIdRaw;
+    const rawText = reply.text ?? "";
+    const text = core.channel.text.convertMarkdownTables(rawText, tableMode);
    const mediaList = reply.mediaUrls?.length
      ? reply.mediaUrls
      : reply.mediaUrl
@@ -43,13 +54,14 @@ export async function deliverMatrixReplies(params: {
      Boolean(id) && (params.replyToMode === "all" || !hasReplied);

    if (mediaList.length === 0) {
-      for (const chunk of core.channel.text.chunkMarkdownText(reply.text ?? "", chunkLimit)) {
+      for (const chunk of core.channel.text.chunkMarkdownText(text, chunkLimit)) {
        const trimmed = chunk.trim();
        if (!trimmed) continue;
        await sendMessageMatrix(params.roomId, trimmed, {
          client: params.client,
          replyToId: shouldIncludeReply(replyToId) ? replyToId : undefined,
          threadId: params.threadId,
+          accountId: params.accountId,
        });
        if (shouldIncludeReply(replyToId)) {
          hasReplied = true;
@@ -60,13 +72,14 @@ export async function deliverMatrixReplies(params: {

    let first = true;
    for (const mediaUrl of mediaList) {
-      const caption = first ? (reply.text ?? "") : "";
+      const caption = first ? text : "";
      await sendMessageMatrix(params.roomId, caption, {
        client: params.client,
        mediaUrl,
        replyToId: shouldIncludeReply(replyToId) ? replyToId : undefined,
        threadId: params.threadId,
        audioAsVoice: reply.audioAsVoice,
+        accountId: params.accountId,
      });
      if (shouldIncludeReply(replyToId)) {
        hasReplied = true;
@@ -43,6 +43,8 @@ const runtimeStub = {
    text: {
      resolveTextChunkLimit: () => 4000,
      chunkMarkdownText: (text: string) => (text ? [text] : []),
+      resolveMarkdownTableMode: () => "code",
+      convertMarkdownTables: (text: string) => text,
    },
  },
 } as unknown as PluginRuntime;
@@ -50,9 +50,18 @@ export async function sendMessageMatrix(
  try {
    const roomId = await resolveMatrixRoomId(client, to);
    const cfg = getCore().config.loadConfig();
+    const tableMode = getCore().channel.text.resolveMarkdownTableMode({
+      cfg,
+      channel: "matrix",
+      accountId: opts.accountId,
+    });
+    const convertedMessage = getCore().channel.text.convertMarkdownTables(
+      trimmedMessage,
+      tableMode,
+    );
    const textLimit = getCore().channel.text.resolveTextChunkLimit(cfg, "matrix");
    const chunkLimit = Math.min(textLimit, MATRIX_TEXT_LIMIT);
-    const chunks = getCore().channel.text.chunkMarkdownText(trimmedMessage, chunkLimit);
+    const chunks = getCore().channel.text.chunkMarkdownText(convertedMessage, chunkLimit);
    const threadId = normalizeThreadId(opts.threadId);
    const relation = threadId
      ? buildThreadRelation(threadId, opts.replyToId)
@@ -0,0 +1,102 @@
+import { beforeEach, describe, expect, it, vi } from "vitest";
+
+import type { MatrixClient } from "matrix-bot-sdk";
+import { EventType } from "./types.js";
+
+let resolveMatrixRoomId: typeof import("./targets.js").resolveMatrixRoomId;
+let normalizeThreadId: typeof import("./targets.js").normalizeThreadId;
+
+beforeEach(async () => {
+  vi.resetModules();
+  ({ resolveMatrixRoomId, normalizeThreadId } = await import("./targets.js"));
+});
+
+describe("resolveMatrixRoomId", () => {
+  it("uses m.direct when available", async () => {
+    const userId = "@user:example.org";
+    const client = {
+      getAccountData: vi.fn().mockResolvedValue({
+        [userId]: ["!room:example.org"],
+      }),
+      getJoinedRooms: vi.fn(),
+      getJoinedRoomMembers: vi.fn(),
+      setAccountData: vi.fn(),
+    } as unknown as MatrixClient;
+
+    const roomId = await resolveMatrixRoomId(client, userId);
+
+    expect(roomId).toBe("!room:example.org");
+    expect(client.getJoinedRooms).not.toHaveBeenCalled();
+    expect(client.setAccountData).not.toHaveBeenCalled();
+  });
+
+  it("falls back to joined rooms and persists m.direct", async () => {
+    const userId = "@fallback:example.org";
+    const roomId = "!room:example.org";
+    const setAccountData = vi.fn().mockResolvedValue(undefined);
+    const client = {
+      getAccountData: vi.fn().mockRejectedValue(new Error("nope")),
+      getJoinedRooms: vi.fn().mockResolvedValue([roomId]),
+      getJoinedRoomMembers: vi.fn().mockResolvedValue([
+        "@bot:example.org",
+        userId,
+      ]),
+      setAccountData,
+    } as unknown as MatrixClient;
+
+    const resolved = await resolveMatrixRoomId(client, userId);
+
+    expect(resolved).toBe(roomId);
+    expect(setAccountData).toHaveBeenCalledWith(
+      EventType.Direct,
+      expect.objectContaining({ [userId]: [roomId] }),
+    );
+  });
+
+  it("continues when a room member lookup fails", async () => {
+    const userId = "@continue:example.org";
+    const roomId = "!good:example.org";
+    const setAccountData = vi.fn().mockResolvedValue(undefined);
+    const getJoinedRoomMembers = vi
+      .fn()
+      .mockRejectedValueOnce(new Error("boom"))
+      .mockResolvedValueOnce(["@bot:example.org", userId]);
+    const client = {
+      getAccountData: vi.fn().mockRejectedValue(new Error("nope")),
+      getJoinedRooms: vi.fn().mockResolvedValue(["!bad:example.org", roomId]),
+      getJoinedRoomMembers,
+      setAccountData,
+    } as unknown as MatrixClient;
+
+    const resolved = await resolveMatrixRoomId(client, userId);
+
+    expect(resolved).toBe(roomId);
+    expect(setAccountData).toHaveBeenCalled();
+  });
+
+  it("allows larger rooms when no 1:1 match exists", async () => {
+    const userId = "@group:example.org";
+    const roomId = "!group:example.org";
+    const client = {
+      getAccountData: vi.fn().mockRejectedValue(new Error("nope")),
+      getJoinedRooms: vi.fn().mockResolvedValue([roomId]),
+      getJoinedRoomMembers: vi.fn().mockResolvedValue([
+        "@bot:example.org",
+        userId,
+        "@extra:example.org",
+      ]),
+      setAccountData: vi.fn().mockResolvedValue(undefined),
+    } as unknown as MatrixClient;
+
+    const resolved = await resolveMatrixRoomId(client, userId);
+
+    expect(resolved).toBe(roomId);
+  });
+});
+
+describe("normalizeThreadId", () => {
+  it("returns null for empty thread ids", () => {
+    expect(normalizeThreadId("   ")).toBeNull();
+    expect(normalizeThreadId("$thread")).toBe("$thread");
+  });
+});
@@ -16,22 +16,100 @@ export function normalizeThreadId(raw?: string | number | null): string | null {
  return trimmed ? trimmed : null;
 }

-async function resolveDirectRoomId(client: MatrixClient, userId: string): Promise<string> {
+const directRoomCache = new Map<string, string>();
+
+async function persistDirectRoom(
+  client: MatrixClient,
+  userId: string,
+  roomId: string,
+): Promise<void> {
+  let directContent: MatrixDirectAccountData | null = null;
+  try {
+    directContent = (await client.getAccountData(
+      EventType.Direct,
+    )) as MatrixDirectAccountData | null;
+  } catch {
+    // Ignore fetch errors and fall back to an empty map.
+  }
+  const existing =
+    directContent && !Array.isArray(directContent) ? directContent : {};
+  const current = Array.isArray(existing[userId]) ? existing[userId] : [];
+  if (current[0] === roomId) return;
+  const next = [roomId, ...current.filter((id) => id !== roomId)];
+  try {
+    await client.setAccountData(EventType.Direct, {
+      ...existing,
+      [userId]: next,
+    });
+  } catch {
+    // Ignore persistence errors.
+  }
+}
+
+async function resolveDirectRoomId(
+  client: MatrixClient,
+  userId: string,
+): Promise<string> {
  const trimmed = userId.trim();
  if (!trimmed.startsWith("@")) {
-    throw new Error(`Matrix user IDs must be fully qualified (got "${trimmed}")`);
+    throw new Error(
+      `Matrix user IDs must be fully qualified (got "${trimmed}")`,
+    );
  }
-  // matrix-bot-sdk: use getAccountData to retrieve m.direct
+
+  const cached = directRoomCache.get(trimmed);
+  if (cached) return cached;
+
+  // 1) Fast path: use account data (m.direct) for *this* logged-in user (the bot).
  try {
-    const directContent = await client.getAccountData(EventType.Direct) as MatrixDirectAccountData | null;
-    const list = Array.isArray(directContent?.[trimmed]) ? directContent[trimmed] : [];
-    if (list.length > 0) return list[0];
+    const directContent = (await client.getAccountData(
+      EventType.Direct,
+    )) as MatrixDirectAccountData | null;
+    const list = Array.isArray(directContent?.[trimmed])
+      ? directContent[trimmed]
+      : [];
+    if (list.length > 0) {
+      directRoomCache.set(trimmed, list[0]);
+      return list[0];
+    }
  } catch {
-    // Ignore errors, try fetching from server
+    // Ignore and fall back.
  }
-  throw new Error(
-    `No m.direct room found for ${trimmed}. Open a DM first so Matrix can set m.direct.`,
-  );
+
+  // 2) Fallback: look for an existing joined room that looks like a 1:1 with the user.
+  // Many clients only maintain m.direct for *their own* account data, so relying on it is brittle.
+  let fallbackRoom: string | null = null;
+  try {
+    const rooms = await client.getJoinedRooms();
+    for (const roomId of rooms) {
+      let members: string[];
+      try {
+        members = await client.getJoinedRoomMembers(roomId);
+      } catch {
+        continue;
+      }
+      if (!members.includes(trimmed)) continue;
+      // Prefer classic 1:1 rooms, but allow larger rooms if requested.
+      if (members.length === 2) {
+        directRoomCache.set(trimmed, roomId);
+        await persistDirectRoom(client, trimmed, roomId);
+        return roomId;
+      }
+      if (!fallbackRoom) {
+        fallbackRoom = roomId;
+      }
+    }
+  } catch {
+    // Ignore and fall back.
+  }
+
+  if (fallbackRoom) {
+    directRoomCache.set(trimmed, fallbackRoom);
+    await persistDirectRoom(client, trimmed, fallbackRoom);
+    return fallbackRoom;
+  }
+
+  throw new Error(`No direct room found for ${trimmed} (m.direct missing)`);
 }

 export async function resolveMatrixRoomId(
@@ -87,6 +87,7 @@ export type MatrixSendResult = {
 export type MatrixSendOpts = {
  client?: import("matrix-bot-sdk").MatrixClient;
  mediaUrl?: string;
+  accountId?: string;
  replyToId?: string;
  threadId?: string | number | null;
  timeoutMs?: number;
@@ -1,6 +1,6 @@
 {
  "name": "@clawdbot/mattermost",
-  "version": "2026.1.20-2",
+  "version": "2026.1.22",
  "type": "module",
  "description": "Clawdbot Mattermost channel plugin",
  "clawdbot": {
@@ -4,6 +4,7 @@ import {
  BlockStreamingCoalesceSchema,
  DmPolicySchema,
  GroupPolicySchema,
+  MarkdownConfigSchema,
  requireOpenAllowFrom,
 } from "clawdbot/plugin-sdk";

@@ -11,6 +12,7 @@ const MattermostAccountSchemaBase = z
  .object({
    name: z.string().optional(),
    capabilities: z.array(z.string()).optional(),
+    markdown: MarkdownConfigSchema,
    enabled: z.boolean().optional(),
    configWrites: z.boolean().optional(),
    botToken: z.string().optional(),
@@ -707,6 +707,11 @@ export async function monitorMattermostProvider(opts: MonitorMattermostOpts = {}
    const textLimit = core.channel.text.resolveTextChunkLimit(cfg, "mattermost", account.accountId, {
      fallbackLimit: account.textChunkLimit ?? 4000,
    });
+    const tableMode = core.channel.text.resolveMarkdownTableMode({
+      cfg,
+      channel: "mattermost",
+      accountId: account.accountId,
+    });

    let prefixContext: ResponsePrefixContext = {
      identityName: resolveIdentityName(cfg, route.agentId),
@@ -720,7 +725,7 @@ export async function monitorMattermostProvider(opts: MonitorMattermostOpts = {}
        humanDelay: core.channel.reply.resolveHumanDelayConfig(cfg, route.agentId),
        deliver: async (payload: ReplyPayload) => {
          const mediaUrls = payload.mediaUrls ?? (payload.mediaUrl ? [payload.mediaUrl] : []);
-          const text = payload.text ?? "";
+          const text = core.channel.text.convertMarkdownTables(payload.text ?? "", tableMode);
          if (mediaUrls.length === 0) {
            const chunks = core.channel.text.chunkMarkdownText(text, textLimit);
            for (const chunk of chunks.length > 0 ? chunks : [text]) {
@@ -181,6 +181,15 @@ export async function sendMessageMattermost(
    }
  }

+  if (message) {
+    const tableMode = core.channel.text.resolveMarkdownTableMode({
+      cfg,
+      channel: "mattermost",
+      accountId: account.accountId,
+    });
+    message = core.channel.text.convertMarkdownTables(message, tableMode);
+  }
+
  if (!message && (!fileIds || fileIds.length === 0)) {
    if (uploadError) {
      throw new Error(`Mattermost media upload failed: ${uploadError.message}`);
@@ -205,4 +214,4 @@ export async function sendMessageMattermost(
    messageId: post.id ?? "unknown",
    channelId,
  };
-}
+}
@@ -14,11 +14,12 @@ import fs from "node:fs/promises";
 import path from "node:path";
 import os from "node:os";

-// Skip if no OpenAI API key
-const OPENAI_API_KEY = process.env.OPENAI_API_KEY;
-const describeWithKey = OPENAI_API_KEY ? describe : describe.skip;
+const OPENAI_API_KEY = process.env.OPENAI_API_KEY ?? "test-key";
+const HAS_OPENAI_KEY = Boolean(process.env.OPENAI_API_KEY);
+const liveEnabled = HAS_OPENAI_KEY && process.env.CLAWDBOT_LIVE_TEST === "1";
+const describeLive = liveEnabled ? describe : describe.skip;

-describeWithKey("memory plugin e2e", () => {
+describe("memory plugin e2e", () => {
  let tmpDir: string;
  let dbPath: string;

@@ -159,7 +160,7 @@ describeWithKey("memory plugin e2e", () => {
 });

 // Live tests that require OpenAI API key and actually use LanceDB
-describeWithKey("memory plugin live tests", () => {
+describeLive("memory plugin live tests", () => {
  let tmpDir: string;
  let dbPath: string;

@@ -176,6 +177,7 @@ describeWithKey("memory plugin live tests", () => {

  test("memory tools work end-to-end", async () => {
    const { default: memoryPlugin } = await import("./index.js");
+    const liveApiKey = process.env.OPENAI_API_KEY ?? "";

    // Mock plugin API
    const registeredTools: any[] = [];
@@ -191,7 +193,7 @@ describeWithKey("memory plugin live tests", () => {
      config: {},
      pluginConfig: {
        embedding: {
-          apiKey: OPENAI_API_KEY,
+          apiKey: liveApiKey,
          model: "text-embedding-3-small",
        },
        dbPath,
@@ -21,6 +21,8 @@ const runtimeStub = {
        }
        return chunks;
      },
+      resolveMarkdownTableMode: () => "code",
+      convertMarkdownTables: (text: string) => text,
    },
  },
 } as unknown as PluginRuntime;
@@ -34,6 +36,7 @@ describe("msteams messenger", () => {
    it("filters silent replies", () => {
      const messages = renderReplyPayloadsToMessages([{ text: SILENT_REPLY_TOKEN }], {
        textChunkLimit: 4000,
+        tableMode: "code",
      });
      expect(messages).toEqual([]);
    });
@@ -41,7 +44,7 @@ describe("msteams messenger", () => {
    it("filters silent reply prefixes", () => {
      const messages = renderReplyPayloadsToMessages(
        [{ text: `${SILENT_REPLY_TOKEN} -- ignored` }],
-        { textChunkLimit: 4000 },
+        { textChunkLimit: 4000, tableMode: "code" },
      );
      expect(messages).toEqual([]);
    });
@@ -49,7 +52,7 @@ describe("msteams messenger", () => {
    it("splits media into separate messages by default", () => {
      const messages = renderReplyPayloadsToMessages(
        [{ text: "hi", mediaUrl: "https://example.com/a.png" }],
-        { textChunkLimit: 4000 },
+        { textChunkLimit: 4000, tableMode: "code" },
      );
      expect(messages).toEqual([{ text: "hi" }, { mediaUrl: "https://example.com/a.png" }]);
    });
@@ -57,7 +60,7 @@ describe("msteams messenger", () => {
    it("supports inline media mode", () => {
      const messages = renderReplyPayloadsToMessages(
        [{ text: "hi", mediaUrl: "https://example.com/a.png" }],
-        { textChunkLimit: 4000, mediaMode: "inline" },
+        { textChunkLimit: 4000, mediaMode: "inline", tableMode: "code" },
      );
      expect(messages).toEqual([{ text: "hi", mediaUrl: "https://example.com/a.png" }]);
    });
@@ -66,6 +69,7 @@ describe("msteams messenger", () => {
      const long = "hello ".repeat(200);
      const messages = renderReplyPayloadsToMessages([{ text: long }], {
        textChunkLimit: 50,
+        tableMode: "code",
      });
      expect(messages.length).toBeGreaterThan(1);
    });
@@ -1,6 +1,7 @@
 import {
  isSilentReplyText,
  loadWebMedia,
+  type MarkdownTableMode,
  type MSTeamsReplyStyle,
  type ReplyPayload,
  SILENT_REPLY_TOKEN,
@@ -61,6 +62,7 @@ export type MSTeamsReplyRenderOptions = {
  textChunkLimit: number;
  chunkText?: boolean;
  mediaMode?: "split" | "inline";
+  tableMode?: MarkdownTableMode;
 };

 /**
@@ -196,10 +198,19 @@ export function renderReplyPayloadsToMessages(
  const chunkLimit = Math.min(options.textChunkLimit, 4000);
  const chunkText = options.chunkText !== false;
  const mediaMode = options.mediaMode ?? "split";
+  const tableMode =
+    options.tableMode ??
+    getMSTeamsRuntime().channel.text.resolveMarkdownTableMode({
+      cfg: getMSTeamsRuntime().config.loadConfig(),
+      channel: "msteams",
+    });

  for (const payload of replies) {
    const mediaList = payload.mediaUrls ?? (payload.mediaUrl ? [payload.mediaUrl] : []);
-    const text = payload.text ?? "";
+    const text = getMSTeamsRuntime().channel.text.convertMarkdownTables(
+      payload.text ?? "",
+      tableMode,
+    );

    if (!text && mediaList.length === 0) continue;

@@ -53,10 +53,15 @@ export function createMSTeamsReplyDispatcher(params: {
    ).responsePrefix,
    humanDelay: core.channel.reply.resolveHumanDelayConfig(params.cfg, params.agentId),
    deliver: async (payload) => {
+      const tableMode = core.channel.text.resolveMarkdownTableMode({
+        cfg: params.cfg,
+        channel: "msteams",
+      });
      const messages = renderReplyPayloadsToMessages([payload], {
        textChunkLimit: params.textLimit,
        chunkText: true,
        mediaMode: "split",
+        tableMode,
      });
      const mediaMaxBytes = resolveChannelMediaMaxBytes({
        cfg: params.cfg,
@@ -16,6 +16,7 @@ import {
 import { extractFilename, extractMessageId } from "./media-helpers.js";
 import { buildConversationReference, sendMSTeamsMessages } from "./messenger.js";
 import { buildMSTeamsPollCard } from "./polls.js";
+import { getMSTeamsRuntime } from "./runtime.js";
 import { resolveMSTeamsSendContext, type MSTeamsProactiveContext } from "./send-context.js";

 export type SendMSTeamsMessageParams = {
@@ -93,13 +94,21 @@ export async function sendMessageMSTeams(
  params: SendMSTeamsMessageParams,
 ): Promise<SendMSTeamsMessageResult> {
  const { cfg, to, text, mediaUrl } = params;
+  const tableMode = getMSTeamsRuntime().channel.text.resolveMarkdownTableMode({
+    cfg,
+    channel: "msteams",
+  });
+  const messageText = getMSTeamsRuntime().channel.text.convertMarkdownTables(
+    text ?? "",
+    tableMode,
+  );
  const ctx = await resolveMSTeamsSendContext({ cfg, to });
  const { adapter, appId, conversationId, ref, log, conversationType, tokenProvider, sharePointSiteId } = ctx;

  log.debug("sending proactive message", {
    conversationId,
    conversationType,
-    textLength: text.length,
+    textLength: messageText.length,
    hasMedia: Boolean(mediaUrl),
  });

@@ -134,7 +143,7 @@ export async function sendMessageMSTeams(
      const { activity, uploadId } = prepareFileConsentActivity({
        media: { buffer: media.buffer, filename: fileName, contentType: media.contentType },
        conversationId,
-        description: text || undefined,
+        description: messageText || undefined,
      });

      log.debug("sending file consent card", { uploadId, fileName, size: media.buffer.length });
@@ -172,14 +181,14 @@ export async function sendMessageMSTeams(
      const base64 = media.buffer.toString("base64");
      const finalMediaUrl = `data:${media.contentType};base64,${base64}`;

-      return sendTextWithMedia(ctx, text, finalMediaUrl);
+      return sendTextWithMedia(ctx, messageText, finalMediaUrl);
    }

    if (isImage && !sharePointSiteId) {
      // Group chat/channel without SharePoint: send image inline (avoids OneDrive failures)
      const base64 = media.buffer.toString("base64");
      const finalMediaUrl = `data:${media.contentType};base64,${base64}`;
-      return sendTextWithMedia(ctx, text, finalMediaUrl);
+      return sendTextWithMedia(ctx, messageText, finalMediaUrl);
    }

    // Group chat or channel: upload to SharePoint (if siteId configured) or OneDrive
@@ -223,7 +232,7 @@ export async function sendMessageMSTeams(
        const fileCardAttachment = buildTeamsFileInfoCard(driveItem);
        const activity = {
          type: "message",
-          text: text || undefined,
+          text: messageText || undefined,
          attachments: [fileCardAttachment],
        };

@@ -264,7 +273,7 @@ export async function sendMessageMSTeams(
      const fileLink = `📎 [${uploaded.name}](${uploaded.shareUrl})`;
      const activity = {
        type: "message",
-        text: text ? `${text}\n\n${fileLink}` : fileLink,
+        text: messageText ? `${messageText}\n\n${fileLink}` : fileLink,
      };

      const baseRef = buildConversationReference(ref);
@@ -290,7 +299,7 @@ export async function sendMessageMSTeams(
  }

  // No media: send text only
-  return sendTextWithMedia(ctx, text, undefined);
+  return sendTextWithMedia(ctx, messageText, undefined);
 }

 /**
@@ -3,6 +3,7 @@ import {
  DmConfigSchema,
  DmPolicySchema,
  GroupPolicySchema,
+  MarkdownConfigSchema,
  requireOpenAllowFrom,
 } from "clawdbot/plugin-sdk";
 import { z } from "zod";
@@ -21,6 +22,7 @@ export const NextcloudTalkAccountSchemaBase = z
  .object({
    name: z.string().optional(),
    enabled: z.boolean().optional(),
+    markdown: MarkdownConfigSchema,
    baseUrl: z.string().optional(),
    botSecret: z.string().optional(),
    botSecretFile: z.string().optional(),
@@ -71,8 +71,18 @@ export async function sendMessageNextcloudTalk(
    throw new Error("Message must be non-empty for Nextcloud Talk sends");
  }

+  const tableMode = getNextcloudTalkRuntime().channel.text.resolveMarkdownTableMode({
+    cfg,
+    channel: "nextcloud-talk",
+    accountId: account.accountId,
+  });
+  const message = getNextcloudTalkRuntime().channel.text.convertMarkdownTables(
+    text.trim(),
+    tableMode,
+  );
+
  const body: Record<string, unknown> = {
-    message: text.trim(),
+    message,
  };
  if (opts.replyTo) {
    body.replyTo = opts.replyTo;
@@ -133,13 +133,20 @@ export const nostrPlugin: ChannelPlugin<ResolvedNostrAccount> = {
    deliveryMode: "direct",
    textChunkLimit: 4000,
    sendText: async ({ to, text, accountId }) => {
+      const core = getNostrRuntime();
      const aid = accountId ?? DEFAULT_ACCOUNT_ID;
      const bus = activeBuses.get(aid);
      if (!bus) {
        throw new Error(`Nostr bus not running for account ${aid}`);
      }
+      const tableMode = core.channel.text.resolveMarkdownTableMode({
+        cfg: core.config.loadConfig(),
+        channel: "nostr",
+        accountId: aid,
+      });
+      const message = core.channel.text.convertMarkdownTables(text ?? "", tableMode);
      const normalizedTo = normalizePubkey(to);
-      await bus.sendDm(normalizedTo, text);
+      await bus.sendDm(normalizedTo, message);
      return { channel: "nostr", to: normalizedTo };
    },
  },
@@ -1,5 +1,5 @@
+import { MarkdownConfigSchema, buildChannelConfigSchema } from "clawdbot/plugin-sdk";
 import { z } from "zod";
-import { buildChannelConfigSchema } from "clawdbot/plugin-sdk";

 const allowFromEntry = z.union([z.string(), z.number()]);

@@ -63,6 +63,9 @@ export const NostrConfigSchema = z.object({
  /** Whether this channel is enabled */
  enabled: z.boolean().optional(),

+  /** Markdown formatting overrides (tables). */
+  markdown: MarkdownConfigSchema,
+
  /** Private key in hex or nsec bech32 format */
  privateKey: z.string().optional(),

@@ -1,6 +1,6 @@
 {
  "name": "@clawdbot/open-prose",
-  "version": "2026.1.23",
+  "version": "2026.1.22",
  "type": "module",
  "description": "OpenProse VM skill pack plugin (slash command + telemetry).",
  "clawdbot": {
@@ -1,3 +1,4 @@
+import { MarkdownConfigSchema } from "clawdbot/plugin-sdk";
 import { z } from "zod";

 const allowFromEntry = z.union([z.string(), z.number()]);
@@ -5,6 +6,7 @@ const allowFromEntry = z.union([z.string(), z.number()]);
 const zaloAccountSchema = z.object({
  name: z.string().optional(),
  enabled: z.boolean().optional(),
+  markdown: MarkdownConfigSchema,
  botToken: z.string().optional(),
  tokenFile: z.string().optional(),
  webhookUrl: z.string().optional(),
@@ -1,6 +1,6 @@
 import type { IncomingMessage, ServerResponse } from "node:http";

-import type { ClawdbotConfig } from "clawdbot/plugin-sdk";
+import type { ClawdbotConfig, MarkdownTableMode } from "clawdbot/plugin-sdk";

 import type { ResolvedZaloAccount } from "./accounts.js";
 import {
@@ -578,6 +578,12 @@ async function processMessageWithPipeline(params: {
    runtime.error?.(`zalo: failed updating session meta: ${String(err)}`);
  });

+  const tableMode = core.channel.text.resolveMarkdownTableMode({
+    cfg: config,
+    channel: "zalo",
+    accountId: account.accountId,
+  });
+
  await core.channel.reply.dispatchReplyWithBufferedBlockDispatcher({
    ctx: ctxPayload,
    cfg: config,
@@ -591,6 +597,7 @@ async function processMessageWithPipeline(params: {
          core,
          statusSink,
          fetcher,
+          tableMode,
        });
      },
      onError: (err, info) => {
@@ -608,8 +615,11 @@ async function deliverZaloReply(params: {
  core: ZaloCoreRuntime;
  statusSink?: (patch: { lastInboundAt?: number; lastOutboundAt?: number }) => void;
  fetcher?: ZaloFetch;
+  tableMode?: MarkdownTableMode;
 }): Promise<void> {
  const { payload, token, chatId, runtime, core, statusSink, fetcher } = params;
+  const tableMode = params.tableMode ?? "code";
+  const text = core.channel.text.convertMarkdownTables(payload.text ?? "", tableMode);

  const mediaList = payload.mediaUrls?.length
    ? payload.mediaUrls
@@ -620,7 +630,7 @@ async function deliverZaloReply(params: {
  if (mediaList.length > 0) {
    let first = true;
    for (const mediaUrl of mediaList) {
-      const caption = first ? payload.text : undefined;
+      const caption = first ? text : undefined;
      first = false;
      try {
        await sendPhoto(token, { chat_id: chatId, photo: mediaUrl, caption }, fetcher);
@@ -632,8 +642,8 @@ async function deliverZaloReply(params: {
    return;
  }

-  if (payload.text) {
-    const chunks = core.channel.text.chunkMarkdownText(payload.text, ZALO_TEXT_LIMIT);
+  if (text) {
+    const chunks = core.channel.text.chunkMarkdownText(text, ZALO_TEXT_LIMIT);
    for (const chunk of chunks) {
      try {
        await sendMessage(token, { chat_id: chatId, text: chunk }, fetcher);
@@ -1,3 +1,4 @@
+import { MarkdownConfigSchema } from "clawdbot/plugin-sdk";
 import { z } from "zod";

 const allowFromEntry = z.union([z.string(), z.number()]);
@@ -10,6 +11,7 @@ const groupConfigSchema = z.object({
 const zalouserAccountSchema = z.object({
  name: z.string().optional(),
  enabled: z.boolean().optional(),
+  markdown: MarkdownConfigSchema,
  profile: z.string().optional(),
  dmPolicy: z.enum(["pairing", "allowlist", "open", "disabled"]).optional(),
  allowFrom: z.array(allowFromEntry).optional(),
@@ -1,6 +1,6 @@
 import type { ChildProcess } from "node:child_process";

-import type { ClawdbotConfig, RuntimeEnv } from "clawdbot/plugin-sdk";
+import type { ClawdbotConfig, MarkdownTableMode, RuntimeEnv } from "clawdbot/plugin-sdk";
 import { mergeAllowlist, summarizeMapping } from "clawdbot/plugin-sdk";
 import { sendMessageZalouser } from "./send.js";
 import type {
@@ -332,6 +332,11 @@ async function processMessage(
          runtime,
          core,
          statusSink,
+          tableMode: core.channel.text.resolveMarkdownTableMode({
+            cfg: config,
+            channel: "zalouser",
+            accountId: account.accountId,
+          }),
        });
      },
      onError: (err, info) => {
@@ -351,8 +356,11 @@ async function deliverZalouserReply(params: {
  runtime: RuntimeEnv;
  core: ZalouserCoreRuntime;
  statusSink?: (patch: { lastInboundAt?: number; lastOutboundAt?: number }) => void;
+  tableMode?: MarkdownTableMode;
 }): Promise<void> {
  const { payload, profile, chatId, isGroup, runtime, core, statusSink } = params;
+  const tableMode = params.tableMode ?? "code";
+  const text = core.channel.text.convertMarkdownTables(payload.text ?? "", tableMode);

  const mediaList = payload.mediaUrls?.length
    ? payload.mediaUrls
@@ -363,7 +371,7 @@ async function deliverZalouserReply(params: {
  if (mediaList.length > 0) {
    let first = true;
    for (const mediaUrl of mediaList) {
-      const caption = first ? payload.text : undefined;
+      const caption = first ? text : undefined;
      first = false;
      try {
        logVerbose(core, runtime, `Sending media to ${chatId}`);
@@ -380,8 +388,8 @@ async function deliverZalouserReply(params: {
    return;
  }

-  if (payload.text) {
-    const chunks = core.channel.text.chunkMarkdownText(payload.text, ZALOUSER_TEXT_LIMIT);
+  if (text) {
+    const chunks = core.channel.text.chunkMarkdownText(text, ZALOUSER_TEXT_LIMIT);
    logVerbose(core, runtime, `Sending ${chunks.length} text chunk(s) to ${chatId}`);
    for (const chunk of chunks) {
      try {
@@ -111,7 +111,7 @@
    "format:swift": "swiftformat --lint --config .swiftformat apps/macos/Sources apps/ios/Sources apps/shared/ClawdbotKit/Sources",
    "format:all": "pnpm format && pnpm format:swift",
    "format:fix": "oxfmt --write src test",
-    "test": "vitest run",
+    "test": "node scripts/test-parallel.mjs",
    "test:watch": "vitest",
    "test:ui": "pnpm --dir ui test",
    "test:force": "node --import tsx scripts/test-force.ts",
@@ -19,7 +19,12 @@ echo "==> Verify git installed"
 command -v git >/dev/null

 echo "==> Verify clawdbot installed"
-LATEST_VERSION="$(npm view clawdbot version)"
+EXPECTED_VERSION="${CLAWDBOT_INSTALL_EXPECT_VERSION:-}"
+if [[ -n "$EXPECTED_VERSION" ]]; then
+  LATEST_VERSION="$EXPECTED_VERSION"
+else
+  LATEST_VERSION="$(npm view clawdbot version)"
+fi
 CMD_PATH="$(command -v clawdbot || true)"
 if [[ -z "$CMD_PATH" && -x "$HOME/.npm-global/bin/clawdbot" ]]; then
  CMD_PATH="$HOME/.npm-global/bin/clawdbot"
@@ -6,21 +6,36 @@ SMOKE_PREVIOUS_VERSION="${CLAWDBOT_INSTALL_SMOKE_PREVIOUS:-}"
 SKIP_PREVIOUS="${CLAWDBOT_INSTALL_SMOKE_SKIP_PREVIOUS:-0}"

 echo "==> Resolve npm versions"
-LATEST_VERSION="$(npm view clawdbot version)"
 if [[ -n "$SMOKE_PREVIOUS_VERSION" ]]; then
+  LATEST_VERSION="$(npm view clawdbot version)"
  PREVIOUS_VERSION="$SMOKE_PREVIOUS_VERSION"
 else
-  PREVIOUS_VERSION="$(node - <<'NODE'
-const { execSync } = require("node:child_process");
-
-const versions = JSON.parse(execSync("npm view clawdbot versions --json", { encoding: "utf8" }));
-if (!Array.isArray(versions) || versions.length === 0) {
+  VERSIONS_JSON="$(npm view clawdbot versions --json)"
+  versions_line="$(node - <<'NODE'
+const raw = process.env.VERSIONS_JSON || "[]";
+let versions;
+try {
+  versions = JSON.parse(raw);
+} catch {
+  versions = raw ? [raw] : [];
+}
+if (!Array.isArray(versions)) {
+  versions = [versions];
+}
+if (versions.length === 0) {
  process.exit(1);
 }
-const previous = versions.length >= 2 ? versions[versions.length - 2] : versions[0];
-process.stdout.write(previous);
+const latest = versions[versions.length - 1];
+const previous = versions.length >= 2 ? versions[versions.length - 2] : latest;
+process.stdout.write(`${latest} ${previous}`);
 NODE
 )"
+  LATEST_VERSION="${versions_line%% *}"
+  PREVIOUS_VERSION="${versions_line#* }"
+fi
+
+if [[ -n "${CLAWDBOT_INSTALL_LATEST_OUT:-}" ]]; then
+  printf "%s" "$LATEST_VERSION" > "$CLAWDBOT_INSTALL_LATEST_OUT"
 fi

 echo "latest=$LATEST_VERSION previous=$PREVIOUS_VERSION"
@@ -6,6 +6,9 @@ SMOKE_IMAGE="${CLAWDBOT_INSTALL_SMOKE_IMAGE:-clawdbot-install-smoke:local}"
 NONROOT_IMAGE="${CLAWDBOT_INSTALL_NONROOT_IMAGE:-clawdbot-install-nonroot:local}"
 INSTALL_URL="${CLAWDBOT_INSTALL_URL:-https://clawd.bot/install.sh}"
 CLI_INSTALL_URL="${CLAWDBOT_INSTALL_CLI_URL:-https://clawd.bot/install-cli.sh}"
+SKIP_NONROOT="${CLAWDBOT_INSTALL_SMOKE_SKIP_NONROOT:-0}"
+LATEST_DIR="$(mktemp -d)"
+LATEST_FILE="${LATEST_DIR}/latest"

 echo "==> Build smoke image (upgrade, root): $SMOKE_IMAGE"
 docker build \
@@ -15,31 +18,48 @@ docker build \

 echo "==> Run installer smoke test (root): $INSTALL_URL"
 docker run --rm -t \
+  -v "${LATEST_DIR}:/out" \
  -e CLAWDBOT_INSTALL_URL="$INSTALL_URL" \
+  -e CLAWDBOT_INSTALL_LATEST_OUT="/out/latest" \
  -e CLAWDBOT_INSTALL_SMOKE_PREVIOUS="${CLAWDBOT_INSTALL_SMOKE_PREVIOUS:-}" \
  -e CLAWDBOT_INSTALL_SMOKE_SKIP_PREVIOUS="${CLAWDBOT_INSTALL_SMOKE_SKIP_PREVIOUS:-0}" \
  -e CLAWDBOT_NO_ONBOARD=1 \
  -e DEBIAN_FRONTEND=noninteractive \
  "$SMOKE_IMAGE"

-echo "==> Build non-root image: $NONROOT_IMAGE"
-docker build \
-  -t "$NONROOT_IMAGE" \
-  -f "$ROOT_DIR/scripts/docker/install-sh-nonroot/Dockerfile" \
-  "$ROOT_DIR/scripts/docker/install-sh-nonroot"
+LATEST_VERSION=""
+if [[ -f "$LATEST_FILE" ]]; then
+  LATEST_VERSION="$(cat "$LATEST_FILE")"
+fi

-echo "==> Run installer non-root test: $INSTALL_URL"
-docker run --rm -t \
-  -e CLAWDBOT_INSTALL_URL="$INSTALL_URL" \
-  -e CLAWDBOT_NO_ONBOARD=1 \
-  -e DEBIAN_FRONTEND=noninteractive \
-  "$NONROOT_IMAGE"
+if [[ "$SKIP_NONROOT" == "1" ]]; then
+  echo "==> Skip non-root installer smoke (CLAWDBOT_INSTALL_SMOKE_SKIP_NONROOT=1)"
+else
+  echo "==> Build non-root image: $NONROOT_IMAGE"
+  docker build \
+    -t "$NONROOT_IMAGE" \
+    -f "$ROOT_DIR/scripts/docker/install-sh-nonroot/Dockerfile" \
+    "$ROOT_DIR/scripts/docker/install-sh-nonroot"
+
+  echo "==> Run installer non-root test: $INSTALL_URL"
+  docker run --rm -t \
+    -e CLAWDBOT_INSTALL_URL="$INSTALL_URL" \
+    -e CLAWDBOT_INSTALL_EXPECT_VERSION="$LATEST_VERSION" \
+    -e CLAWDBOT_NO_ONBOARD=1 \
+    -e DEBIAN_FRONTEND=noninteractive \
+    "$NONROOT_IMAGE"
+fi

 if [[ "${CLAWDBOT_INSTALL_SMOKE_SKIP_CLI:-0}" == "1" ]]; then
  echo "==> Skip CLI installer smoke (CLAWDBOT_INSTALL_SMOKE_SKIP_CLI=1)"
  exit 0
 fi

+if [[ "$SKIP_NONROOT" == "1" ]]; then
+  echo "==> Skip CLI installer smoke (non-root image skipped)"
+  exit 0
+fi
+
 echo "==> Run CLI installer non-root test (same image)"
 docker run --rm -t \
  --entrypoint /bin/bash \
@@ -0,0 +1,70 @@
+import { spawn } from "node:child_process";
+import os from "node:os";
+
+const pnpm = process.platform === "win32" ? "pnpm.cmd" : "pnpm";
+
+const runs = [
+  {
+    name: "unit",
+    args: ["vitest", "run", "--config", "vitest.unit.config.ts"],
+  },
+  {
+    name: "extensions",
+    args: ["vitest", "run", "--config", "vitest.extensions.config.ts"],
+  },
+  {
+    name: "gateway",
+    args: ["vitest", "run", "--config", "vitest.gateway.config.ts"],
+  },
+];
+
+const parallelRuns = runs.filter((entry) => entry.name !== "gateway");
+const serialRuns = runs.filter((entry) => entry.name === "gateway");
+
+const children = new Set();
+const isCI = process.env.CI === "true" || process.env.GITHUB_ACTIONS === "true";
+const overrideWorkers = Number.parseInt(process.env.CLAWDBOT_TEST_WORKERS ?? "", 10);
+const resolvedOverride = Number.isFinite(overrideWorkers) && overrideWorkers > 0 ? overrideWorkers : null;
+const localWorkers = Math.max(4, Math.min(16, os.cpus().length));
+const perRunWorkers = Math.max(1, Math.floor(localWorkers / parallelRuns.length));
+const maxWorkers = isCI ? null : resolvedOverride ?? perRunWorkers;
+
+const run = (entry) =>
+  new Promise((resolve) => {
+    const args = maxWorkers ? [...entry.args, "--maxWorkers", String(maxWorkers)] : entry.args;
+    const child = spawn(pnpm, args, {
+      stdio: "inherit",
+      env: { ...process.env, VITEST_GROUP: entry.name },
+      shell: process.platform === "win32",
+    });
+    children.add(child);
+    child.on("exit", (code, signal) => {
+      children.delete(child);
+      resolve(code ?? (signal ? 1 : 0));
+    });
+  });
+
+const shutdown = (signal) => {
+  for (const child of children) {
+    child.kill(signal);
+  }
+};
+
+process.on("SIGINT", () => shutdown("SIGINT"));
+process.on("SIGTERM", () => shutdown("SIGTERM"));
+
+const parallelCodes = await Promise.all(parallelRuns.map(run));
+const failedParallel = parallelCodes.find((code) => code !== 0);
+if (failedParallel !== undefined) {
+  process.exit(failedParallel);
+}
+
+for (const entry of serialRuns) {
+  // eslint-disable-next-line no-await-in-loop
+  const code = await run(entry);
+  if (code !== 0) {
+    process.exit(code);
+  }
+}
+
+process.exit(0);
@@ -1,70 +0,0 @@
-import fs from "node:fs/promises";
-import os from "node:os";
-import path from "node:path";
-
-import { afterEach, describe, expect, it, vi } from "vitest";
-import {
-  type AuthProfileStore,
-  ensureAuthProfileStore,
-  resolveApiKeyForProfile,
-} from "./auth-profiles.js";
-
-vi.mock("@mariozechner/pi-ai", () => ({
-  getOAuthApiKey: vi.fn(() => {
-    throw new Error("refresh should not be called");
-  }),
-}));
-
-describe("auth-profiles (github-copilot)", () => {
-  const previousStateDir = process.env.CLAWDBOT_STATE_DIR;
-  const previousAgentDir = process.env.CLAWDBOT_AGENT_DIR;
-  const previousPiAgentDir = process.env.PI_CODING_AGENT_DIR;
-  let tempDir: string | null = null;
-
-  afterEach(async () => {
-    vi.unstubAllGlobals();
-    if (tempDir) {
-      await fs.rm(tempDir, { recursive: true, force: true });
-      tempDir = null;
-    }
-    if (previousStateDir === undefined) delete process.env.CLAWDBOT_STATE_DIR;
-    else process.env.CLAWDBOT_STATE_DIR = previousStateDir;
-    if (previousAgentDir === undefined) delete process.env.CLAWDBOT_AGENT_DIR;
-    else process.env.CLAWDBOT_AGENT_DIR = previousAgentDir;
-    if (previousPiAgentDir === undefined) delete process.env.PI_CODING_AGENT_DIR;
-    else process.env.PI_CODING_AGENT_DIR = previousPiAgentDir;
-  });
-
-  it("treats copilot oauth tokens with expires=0 as non-expiring", async () => {
-    tempDir = await fs.mkdtemp(path.join(os.tmpdir(), "clawdbot-copilot-"));
-    process.env.CLAWDBOT_STATE_DIR = tempDir;
-    process.env.CLAWDBOT_AGENT_DIR = path.join(tempDir, "agents", "main", "agent");
-    process.env.PI_CODING_AGENT_DIR = process.env.CLAWDBOT_AGENT_DIR;
-
-    const authProfilePath = path.join(tempDir, "agents", "main", "agent", "auth-profiles.json");
-    await fs.mkdir(path.dirname(authProfilePath), { recursive: true });
-
-    const store: AuthProfileStore = {
-      version: 1,
-      profiles: {
-        "github-copilot:github": {
-          type: "oauth",
-          provider: "github-copilot",
-          refresh: "gh-token",
-          access: "gh-token",
-          expires: 0,
-          enterpriseUrl: "company.ghe.com",
-        },
-      },
-    };
-    await fs.writeFile(authProfilePath, `${JSON.stringify(store)}\n`);
-
-    const loaded = ensureAuthProfileStore();
-    const resolved = await resolveApiKeyForProfile({
-      store: loaded,
-      profileId: "github-copilot:github",
-    });
-
-    expect(resolved?.apiKey).toBe("gh-token");
-  });
-});
@@ -39,15 +39,6 @@ async function refreshOAuthTokenWithLock(params: {
    const cred = store.profiles[params.profileId];
    if (!cred || cred.type !== "oauth") return null;

-    if (
-      cred.provider === "github-copilot" &&
-      (!Number.isFinite(cred.expires) || cred.expires <= 0)
-    ) {
-      return {
-        apiKey: buildOAuthApiKey(cred.provider, cred),
-        newCredentials: cred,
-      };
-    }
    if (Date.now() < cred.expires) {
      return {
        apiKey: buildOAuthApiKey(cred.provider, cred),
@@ -112,20 +103,6 @@ async function tryResolveOAuthProfile(params: {
  if (profileConfig && profileConfig.provider !== cred.provider) return null;
  if (profileConfig && profileConfig.mode !== cred.type) return null;

-  if (cred.provider === "github-copilot" && (!Number.isFinite(cred.expires) || cred.expires <= 0)) {
-    return {
-      apiKey: buildOAuthApiKey(cred.provider, cred),
-      provider: cred.provider,
-      email: cred.email,
-    };
-  }
-  if (cred.provider === "github-copilot" && (!Number.isFinite(cred.expires) || cred.expires <= 0)) {
-    return {
-      apiKey: buildOAuthApiKey(cred.provider, cred),
-      provider: cred.provider,
-      email: cred.email,
-    };
-  }
  if (Date.now() < cred.expires) {
    return {
      apiKey: buildOAuthApiKey(cred.provider, cred),
@@ -19,7 +19,6 @@ export type TokenCredential = {
  token: string;
  /** Optional expiry timestamp (ms since epoch). */
  expires?: number;
-  enterpriseUrl?: string;
  email?: string;
 };

@@ -0,0 +1,31 @@
+import { afterEach, expect, test, vi } from "vitest";
+
+import { resetProcessRegistryForTests } from "./bash-process-registry";
+
+afterEach(() => {
+  resetProcessRegistryForTests();
+  vi.resetModules();
+  vi.clearAllMocks();
+});
+
+test("exec falls back when PTY spawn fails", async () => {
+  vi.doMock("@lydell/node-pty", () => ({
+    spawn: () => {
+      const err = new Error("spawn EBADF");
+      (err as NodeJS.ErrnoException).code = "EBADF";
+      throw err;
+    },
+  }));
+
+  const { createExecTool } = await import("./bash-tools.exec");
+  const tool = createExecTool({ allowBackground: false });
+  const result = await tool.execute("toolcall", {
+    command: "printf ok",
+    pty: true,
+  });
+
+  expect(result.details.status).toBe("completed");
+  const text = result.content?.[0]?.text ?? "";
+  expect(text).toContain("ok");
+  expect(text).toContain("PTY spawn failed");
+});
@@ -26,7 +26,7 @@ import {
  resolveShellEnvFallbackTimeoutMs,
 } from "../infra/shell-env.js";
 import { enqueueSystemEvent } from "../infra/system-events.js";
-import { logInfo } from "../logger.js";
+import { logInfo, logWarn } from "../logger.js";
 import {
  type ProcessSession,
  type SessionStdin,
@@ -381,41 +381,56 @@ async function runExecProcess(opts: {
    ) as ChildProcessWithoutNullStreams;
    stdin = child.stdin;
  } else if (opts.usePty) {
-    const ptyModule = (await import("@lydell/node-pty")) as unknown as {
-      spawn?: PtySpawn;
-      default?: { spawn?: PtySpawn };
-    };
-    const spawnPty = ptyModule.spawn ?? ptyModule.default?.spawn;
-    if (!spawnPty) {
-      throw new Error("PTY support is unavailable (node-pty spawn not found).");
-    }
    const { shell, args: shellArgs } = getShellConfig();
-    pty = spawnPty(shell, [...shellArgs, opts.command], {
-      cwd: opts.workdir,
-      env: opts.env,
-      name: process.env.TERM ?? "xterm-256color",
-      cols: 120,
-      rows: 30,
-    });
-    stdin = {
-      destroyed: false,
-      write: (data, cb) => {
-        try {
-          pty?.write(data);
-          cb?.(null);
-        } catch (err) {
-          cb?.(err as Error);
-        }
-      },
-      end: () => {
-        try {
-          const eof = process.platform === "win32" ? "\x1a" : "\x04";
-          pty?.write(eof);
-        } catch {
-          // ignore EOF errors
-        }
-      },
-    };
+    try {
+      const ptyModule = (await import("@lydell/node-pty")) as unknown as {
+        spawn?: PtySpawn;
+        default?: { spawn?: PtySpawn };
+      };
+      const spawnPty = ptyModule.spawn ?? ptyModule.default?.spawn;
+      if (!spawnPty) {
+        throw new Error("PTY support is unavailable (node-pty spawn not found).");
+      }
+      pty = spawnPty(shell, [...shellArgs, opts.command], {
+        cwd: opts.workdir,
+        env: opts.env,
+        name: process.env.TERM ?? "xterm-256color",
+        cols: 120,
+        rows: 30,
+      });
+      stdin = {
+        destroyed: false,
+        write: (data, cb) => {
+          try {
+            pty?.write(data);
+            cb?.(null);
+          } catch (err) {
+            cb?.(err as Error);
+          }
+        },
+        end: () => {
+          try {
+            const eof = process.platform === "win32" ? "\x1a" : "\x04";
+            pty?.write(eof);
+          } catch {
+            // ignore EOF errors
+          }
+        },
+      };
+    } catch (err) {
+      const errText = String(err);
+      const warning = `Warning: PTY spawn failed (${errText}); retrying without PTY for \`${opts.command}\`.`;
+      logWarn(`exec: PTY spawn failed (${errText}); retrying without PTY for "${opts.command}".`);
+      opts.warnings.push(warning);
+      child = spawn(shell, [...shellArgs, opts.command], {
+        cwd: opts.workdir,
+        env: opts.env,
+        detached: process.platform !== "win32",
+        stdio: ["pipe", "pipe", "pipe"],
+        windowsHide: true,
+      }) as ChildProcessWithoutNullStreams;
+      stdin = child.stdin;
+    }
  } else {
    const { shell, args: shellArgs } = getShellConfig();
    child = spawn(shell, [...shellArgs, opts.command], {
@@ -1320,7 +1335,7 @@ export function createExecTool(

      const effectiveTimeout =
        typeof params.timeout === "number" ? params.timeout : defaultTimeoutSec;
-      const warningText = warnings.length ? `${warnings.join("\n")}\n\n` : "";
+      const getWarningText = () => (warnings.length ? `${warnings.join("\n")}\n\n` : "");
      const usePty = params.pty === true && !sandbox;
      const run = await runExecProcess({
        command: params.command,
@@ -1360,7 +1375,7 @@ export function createExecTool(
              {
                type: "text",
                text:
-                  `${warningText}` +
+                  `${getWarningText()}` +
                  `Command still running (session ${run.session.id}, pid ${
                    run.session.pid ?? "n/a"
                  }). ` +
@@ -1410,7 +1425,7 @@ export function createExecTool(
              content: [
                {
                  type: "text",
-                  text: `${warningText}${outcome.aggregated || "(no output)"}`,
+                  text: `${getWarningText()}${outcome.aggregated || "(no output)"}`,
                },
              ],
              details: {
@@ -3,6 +3,7 @@ import os from "node:os";
 import path from "node:path";
 import { describe, expect, it, vi } from "vitest";

+import "./test-helpers/fast-core-tools.js";
 import { createClawdbotTools } from "./clawdbot-tools.js";

 vi.mock("./tools/gateway.js", () => ({
@@ -16,6 +16,7 @@ vi.mock("../config/config.js", async (importOriginal) => {
  };
 });

+import "./test-helpers/fast-core-tools.js";
 import { createClawdbotTools } from "./clawdbot-tools.js";

 describe("agents_list", () => {
@@ -10,6 +10,7 @@ vi.mock("../media/image-ops.js", () => ({
  resizeToJpeg: vi.fn(async () => Buffer.from("jpeg")),
 }));

+import "./test-helpers/fast-core-tools.js";
 import { createClawdbotTools } from "./clawdbot-tools.js";

 describe("nodes camera_snap", () => {
@@ -75,6 +75,7 @@ vi.mock("../infra/provider-usage.js", () => ({
  formatUsageSummaryLine: () => null,
 }));

+import "./test-helpers/fast-core-tools.js";
 import { createClawdbotTools } from "./clawdbot-tools.js";

 describe("session_status tool", () => {
@@ -4,9 +4,6 @@ const callGatewayMock = vi.fn();
 vi.mock("../gateway/call.js", () => ({
  callGateway: (opts: unknown) => callGatewayMock(opts),
 }));
-vi.mock("../plugins/tools.js", () => ({
-  resolvePluginTools: () => [],
-}));

 vi.mock("../config/config.js", async (importOriginal) => {
  const actual = await importOriginal<typeof import("../config/config.js")>();
@@ -23,6 +20,7 @@ vi.mock("../config/config.js", async (importOriginal) => {
  };
 });

+import "./test-helpers/fast-core-tools.js";
 import { createClawdbotTools } from "./clawdbot-tools.js";

 const waitForCalls = async (getCount: () => number, count: number, timeoutMs = 2000) => {
@@ -21,6 +21,7 @@ vi.mock("../config/config.js", async (importOriginal) => {
  };
 });

+import "./test-helpers/fast-core-tools.js";
 import { createClawdbotTools } from "./clawdbot-tools.js";
 import { resetSubagentRegistryForTests } from "./subagent-registry.js";

@@ -21,6 +21,7 @@ vi.mock("../config/config.js", async (importOriginal) => {
  };
 });

+import "./test-helpers/fast-core-tools.js";
 import { createClawdbotTools } from "./clawdbot-tools.js";
 import { resetSubagentRegistryForTests } from "./subagent-registry.js";

@@ -1,235 +0,0 @@
-import { beforeEach, describe, expect, it, vi } from "vitest";
-
-const callGatewayMock = vi.fn();
-vi.mock("../gateway/call.js", () => ({
-  callGateway: (opts: unknown) => callGatewayMock(opts),
-}));
-
-let configOverride: ReturnType<(typeof import("../config/config.js"))["loadConfig"]> = {
-  session: {
-    mainKey: "main",
-    scope: "per-sender",
-  },
-};
-
-vi.mock("../config/config.js", async (importOriginal) => {
-  const actual = await importOriginal<typeof import("../config/config.js")>();
-  return {
-    ...actual,
-    loadConfig: () => configOverride,
-    resolveGatewayPort: () => 18789,
-  };
-});
-
-import { emitAgentEvent } from "../infra/agent-events.js";
-import { createClawdbotTools } from "./clawdbot-tools.js";
-import { resetSubagentRegistryForTests } from "./subagent-registry.js";
-
-describe("clawdbot-tools: subagents", () => {
-  beforeEach(() => {
-    configOverride = {
-      session: {
-        mainKey: "main",
-        scope: "per-sender",
-      },
-    };
-  });
-
-  it("sessions_spawn runs cleanup via lifecycle events", async () => {
-    resetSubagentRegistryForTests();
-    callGatewayMock.mockReset();
-    const calls: Array<{ method?: string; params?: unknown }> = [];
-    let agentCallCount = 0;
-    let deletedKey: string | undefined;
-    let childRunId: string | undefined;
-    let childSessionKey: string | undefined;
-    const waitCalls: Array<{ runId?: string; timeoutMs?: number }> = [];
-
-    callGatewayMock.mockImplementation(async (opts: unknown) => {
-      const request = opts as { method?: string; params?: unknown };
-      calls.push(request);
-      if (request.method === "agent") {
-        agentCallCount += 1;
-        const runId = `run-${agentCallCount}`;
-        const params = request.params as {
-          message?: string;
-          sessionKey?: string;
-          channel?: string;
-          timeout?: number;
-          lane?: string;
-        };
-        // Only capture the first agent call (subagent spawn, not main agent trigger)
-        if (params?.lane === "subagent") {
-          childRunId = runId;
-          childSessionKey = params?.sessionKey ?? "";
-          expect(params?.channel).toBe("discord");
-          expect(params?.timeout).toBe(1);
-        }
-        return {
-          runId,
-          status: "accepted",
-          acceptedAt: 1000 + agentCallCount,
-        };
-      }
-      if (request.method === "agent.wait") {
-        const params = request.params as { runId?: string; timeoutMs?: number } | undefined;
-        waitCalls.push(params ?? {});
-        // Return "ok" with timing info for the child run
-        return { runId: params?.runId ?? "run-1", status: "ok", startedAt: 1000, endedAt: 2000 };
-      }
-      if (request.method === "sessions.delete") {
-        const params = request.params as { key?: string } | undefined;
-        deletedKey = params?.key;
-        return { ok: true };
-      }
-      return {};
-    });
-
-    const tool = createClawdbotTools({
-      agentSessionKey: "discord:group:req",
-      agentChannel: "discord",
-    }).find((candidate) => candidate.name === "sessions_spawn");
-    if (!tool) throw new Error("missing sessions_spawn tool");
-
-    const result = await tool.execute("call1", {
-      task: "do thing",
-      runTimeoutSeconds: 1,
-      cleanup: "delete",
-    });
-    expect(result.details).toMatchObject({
-      status: "accepted",
-      runId: "run-1",
-    });
-
-    if (!childRunId) throw new Error("missing child runId");
-    emitAgentEvent({
-      runId: childRunId,
-      stream: "lifecycle",
-      data: {
-        phase: "end",
-        startedAt: 1234,
-        endedAt: 2345,
-      },
-    });
-
-    await new Promise((resolve) => setTimeout(resolve, 0));
-    await new Promise((resolve) => setTimeout(resolve, 0));
-    await new Promise((resolve) => setTimeout(resolve, 0));
-
-    const childWait = waitCalls.find((call) => call.runId === childRunId);
-    expect(childWait?.timeoutMs).toBe(1000);
-
-    // Two agent calls: subagent spawn + main agent trigger
-    const agentCalls = calls.filter((call) => call.method === "agent");
-    expect(agentCalls).toHaveLength(2);
-
-    // First call: subagent spawn
-    const first = agentCalls[0]?.params as
-      | {
-          lane?: string;
-          deliver?: boolean;
-          sessionKey?: string;
-          channel?: string;
-        }
-      | undefined;
-    expect(first?.lane).toBe("subagent");
-    expect(first?.deliver).toBe(false);
-    expect(first?.channel).toBe("discord");
-    expect(first?.sessionKey?.startsWith("agent:main:subagent:")).toBe(true);
-    expect(childSessionKey?.startsWith("agent:main:subagent:")).toBe(true);
-
-    // Second call: main agent trigger with announce message
-    const second = agentCalls[1]?.params as
-      | {
-          sessionKey?: string;
-          message?: string;
-          deliver?: boolean;
-        }
-      | undefined;
-    expect(second?.sessionKey).toBe("discord:group:req");
-    expect(second?.deliver).toBe(true);
-    expect(second?.message).toContain("background task");
-
-    // No direct send to external channel (main agent handles delivery)
-    const sendCalls = calls.filter((c) => c.method === "send");
-    expect(sendCalls.length).toBe(0);
-
-    // Session should be deleted since cleanup=delete
-    expect(deletedKey?.startsWith("agent:main:subagent:")).toBe(true);
-  });
-
-  it("sessions_spawn announces with requester accountId", async () => {
-    resetSubagentRegistryForTests();
-    callGatewayMock.mockReset();
-    const calls: Array<{ method?: string; params?: unknown }> = [];
-    let agentCallCount = 0;
-    let childRunId: string | undefined;
-
-    callGatewayMock.mockImplementation(async (opts: unknown) => {
-      const request = opts as { method?: string; params?: unknown };
-      calls.push(request);
-      if (request.method === "agent") {
-        agentCallCount += 1;
-        const runId = `run-${agentCallCount}`;
-        const params = request.params as { lane?: string; sessionKey?: string } | undefined;
-        if (params?.lane === "subagent") {
-          childRunId = runId;
-        }
-        return {
-          runId,
-          status: "accepted",
-          acceptedAt: 4000 + agentCallCount,
-        };
-      }
-      if (request.method === "agent.wait") {
-        const params = request.params as { runId?: string; timeoutMs?: number } | undefined;
-        return { runId: params?.runId ?? "run-1", status: "ok", startedAt: 1000, endedAt: 2000 };
-      }
-      if (request.method === "sessions.delete" || request.method === "sessions.patch") {
-        return { ok: true };
-      }
-      return {};
-    });
-
-    const tool = createClawdbotTools({
-      agentSessionKey: "main",
-      agentChannel: "whatsapp",
-      agentAccountId: "kev",
-    }).find((candidate) => candidate.name === "sessions_spawn");
-    if (!tool) throw new Error("missing sessions_spawn tool");
-
-    const result = await tool.execute("call2", {
-      task: "do thing",
-      runTimeoutSeconds: 1,
-      cleanup: "keep",
-    });
-    expect(result.details).toMatchObject({
-      status: "accepted",
-      runId: "run-1",
-    });
-
-    if (!childRunId) throw new Error("missing child runId");
-    emitAgentEvent({
-      runId: childRunId,
-      stream: "lifecycle",
-      data: {
-        phase: "end",
-        startedAt: 1000,
-        endedAt: 2000,
-      },
-    });
-
-    await new Promise((resolve) => setTimeout(resolve, 0));
-    await new Promise((resolve) => setTimeout(resolve, 0));
-    await new Promise((resolve) => setTimeout(resolve, 0));
-
-    const agentCalls = calls.filter((call) => call.method === "agent");
-    expect(agentCalls).toHaveLength(2);
-    const announceParams = agentCalls[1]?.params as
-      | { accountId?: string; channel?: string; deliver?: boolean }
-      | undefined;
-    expect(announceParams?.deliver).toBe(true);
-    expect(announceParams?.channel).toBe("whatsapp");
-    expect(announceParams?.accountId).toBe("kev");
-  });
-});
@@ -21,6 +21,7 @@ vi.mock("../config/config.js", async (importOriginal) => {
  };
 });

+import "./test-helpers/fast-core-tools.js";
 import { createClawdbotTools } from "./clawdbot-tools.js";
 import { resetSubagentRegistryForTests } from "./subagent-registry.js";

@@ -21,6 +21,8 @@ vi.mock("../config/config.js", async (importOriginal) => {
  };
 });

+import { emitAgentEvent } from "../infra/agent-events.js";
+import "./test-helpers/fast-core-tools.js";
 import { createClawdbotTools } from "./clawdbot-tools.js";
 import { resetSubagentRegistryForTests } from "./subagent-registry.js";

@@ -120,4 +122,195 @@ describe("clawdbot-tools: subagents", () => {
    });
    expect(callGatewayMock).not.toHaveBeenCalled();
  });
+
+  it("sessions_spawn runs cleanup via lifecycle events", async () => {
+    resetSubagentRegistryForTests();
+    callGatewayMock.mockReset();
+    const calls: Array<{ method?: string; params?: unknown }> = [];
+    let agentCallCount = 0;
+    let deletedKey: string | undefined;
+    let childRunId: string | undefined;
+    let childSessionKey: string | undefined;
+    const waitCalls: Array<{ runId?: string; timeoutMs?: number }> = [];
+
+    callGatewayMock.mockImplementation(async (opts: unknown) => {
+      const request = opts as { method?: string; params?: unknown };
+      calls.push(request);
+      if (request.method === "agent") {
+        agentCallCount += 1;
+        const runId = `run-${agentCallCount}`;
+        const params = request.params as {
+          message?: string;
+          sessionKey?: string;
+          channel?: string;
+          timeout?: number;
+          lane?: string;
+        };
+        if (params?.lane === "subagent") {
+          childRunId = runId;
+          childSessionKey = params?.sessionKey ?? "";
+          expect(params?.channel).toBe("discord");
+          expect(params?.timeout).toBe(1);
+        }
+        return {
+          runId,
+          status: "accepted",
+          acceptedAt: 1000 + agentCallCount,
+        };
+      }
+      if (request.method === "agent.wait") {
+        const params = request.params as { runId?: string; timeoutMs?: number } | undefined;
+        waitCalls.push(params ?? {});
+        return { runId: params?.runId ?? "run-1", status: "ok", startedAt: 1000, endedAt: 2000 };
+      }
+      if (request.method === "sessions.delete") {
+        const params = request.params as { key?: string } | undefined;
+        deletedKey = params?.key;
+        return { ok: true };
+      }
+      return {};
+    });
+
+    const tool = createClawdbotTools({
+      agentSessionKey: "discord:group:req",
+      agentChannel: "discord",
+    }).find((candidate) => candidate.name === "sessions_spawn");
+    if (!tool) throw new Error("missing sessions_spawn tool");
+
+    const result = await tool.execute("call1", {
+      task: "do thing",
+      runTimeoutSeconds: 1,
+      cleanup: "delete",
+    });
+    expect(result.details).toMatchObject({
+      status: "accepted",
+      runId: "run-1",
+    });
+
+    if (!childRunId) throw new Error("missing child runId");
+    emitAgentEvent({
+      runId: childRunId,
+      stream: "lifecycle",
+      data: {
+        phase: "end",
+        startedAt: 1234,
+        endedAt: 2345,
+      },
+    });
+
+    await new Promise((resolve) => setTimeout(resolve, 0));
+    await new Promise((resolve) => setTimeout(resolve, 0));
+    await new Promise((resolve) => setTimeout(resolve, 0));
+
+    const childWait = waitCalls.find((call) => call.runId === childRunId);
+    expect(childWait?.timeoutMs).toBe(1000);
+
+    const agentCalls = calls.filter((call) => call.method === "agent");
+    expect(agentCalls).toHaveLength(2);
+
+    const first = agentCalls[0]?.params as
+      | {
+          lane?: string;
+          deliver?: boolean;
+          sessionKey?: string;
+          channel?: string;
+        }
+      | undefined;
+    expect(first?.lane).toBe("subagent");
+    expect(first?.deliver).toBe(false);
+    expect(first?.channel).toBe("discord");
+    expect(first?.sessionKey?.startsWith("agent:main:subagent:")).toBe(true);
+    expect(childSessionKey?.startsWith("agent:main:subagent:")).toBe(true);
+
+    const second = agentCalls[1]?.params as
+      | {
+          sessionKey?: string;
+          message?: string;
+          deliver?: boolean;
+        }
+      | undefined;
+    expect(second?.sessionKey).toBe("discord:group:req");
+    expect(second?.deliver).toBe(true);
+    expect(second?.message).toContain("background task");
+
+    const sendCalls = calls.filter((c) => c.method === "send");
+    expect(sendCalls.length).toBe(0);
+
+    expect(deletedKey?.startsWith("agent:main:subagent:")).toBe(true);
+  });
+
+  it("sessions_spawn announces with requester accountId", async () => {
+    resetSubagentRegistryForTests();
+    callGatewayMock.mockReset();
+    const calls: Array<{ method?: string; params?: unknown }> = [];
+    let agentCallCount = 0;
+    let childRunId: string | undefined;
+
+    callGatewayMock.mockImplementation(async (opts: unknown) => {
+      const request = opts as { method?: string; params?: unknown };
+      calls.push(request);
+      if (request.method === "agent") {
+        agentCallCount += 1;
+        const runId = `run-${agentCallCount}`;
+        const params = request.params as { lane?: string; sessionKey?: string } | undefined;
+        if (params?.lane === "subagent") {
+          childRunId = runId;
+        }
+        return {
+          runId,
+          status: "accepted",
+          acceptedAt: 4000 + agentCallCount,
+        };
+      }
+      if (request.method === "agent.wait") {
+        const params = request.params as { runId?: string; timeoutMs?: number } | undefined;
+        return { runId: params?.runId ?? "run-1", status: "ok", startedAt: 1000, endedAt: 2000 };
+      }
+      if (request.method === "sessions.delete" || request.method === "sessions.patch") {
+        return { ok: true };
+      }
+      return {};
+    });
+
+    const tool = createClawdbotTools({
+      agentSessionKey: "main",
+      agentChannel: "whatsapp",
+      agentAccountId: "kev",
+    }).find((candidate) => candidate.name === "sessions_spawn");
+    if (!tool) throw new Error("missing sessions_spawn tool");
+
+    const result = await tool.execute("call2", {
+      task: "do thing",
+      runTimeoutSeconds: 1,
+      cleanup: "keep",
+    });
+    expect(result.details).toMatchObject({
+      status: "accepted",
+      runId: "run-1",
+    });
+
+    if (!childRunId) throw new Error("missing child runId");
+    emitAgentEvent({
+      runId: childRunId,
+      stream: "lifecycle",
+      data: {
+        phase: "end",
+        startedAt: 1000,
+        endedAt: 2000,
+      },
+    });
+
+    await new Promise((resolve) => setTimeout(resolve, 0));
+    await new Promise((resolve) => setTimeout(resolve, 0));
+    await new Promise((resolve) => setTimeout(resolve, 0));
+
+    const agentCalls = calls.filter((call) => call.method === "agent");
+    expect(agentCalls).toHaveLength(2);
+    const announceParams = agentCalls[1]?.params as
+      | { accountId?: string; channel?: string; deliver?: boolean }
+      | undefined;
+    expect(announceParams?.deliver).toBe(true);
+    expect(announceParams?.channel).toBe("whatsapp");
+    expect(announceParams?.accountId).toBe("kev");
+  });
 });
@@ -21,6 +21,7 @@ vi.mock("../config/config.js", async (importOriginal) => {
  };
 });

+import "./test-helpers/fast-core-tools.js";
 import { createClawdbotTools } from "./clawdbot-tools.js";
 import { resetSubagentRegistryForTests } from "./subagent-registry.js";

@@ -22,6 +22,7 @@ vi.mock("../config/config.js", async (importOriginal) => {
 });

 import { emitAgentEvent } from "../infra/agent-events.js";
+import "./test-helpers/fast-core-tools.js";
 import { createClawdbotTools } from "./clawdbot-tools.js";
 import { resetSubagentRegistryForTests } from "./subagent-registry.js";

@@ -66,6 +66,10 @@ export function isModernModelRef(ref: ModelRef): boolean {
    return matchesPrefix(id, XAI_PREFIXES);
  }

+  if (provider === "opencode" && id.endsWith("-free")) {
+    return false;
+  }
+
  if (provider === "openrouter" || provider === "opencode") {
    return matchesAny(id, [
      ...ANTHROPIC_PREFIXES,
@@ -3,7 +3,6 @@ import path from "node:path";
 import { afterEach, beforeEach, describe, expect, it, vi } from "vitest";
 import { withTempHome as withTempHomeBase } from "../../test/helpers/temp-home.js";
 import type { ClawdbotConfig } from "../config/config.js";
-import { DEFAULT_GITHUB_COPILOT_BASE_URL } from "../providers/github-copilot-utils.js";

 async function withTempHome<T>(fn: (home: string) => Promise<T>): Promise<T> {
  return withTempHomeBase(fn, { prefix: "clawdbot-models-" });
@@ -52,6 +51,16 @@ describe("models-config", () => {
      try {
        vi.resetModules();

+        vi.doMock("../providers/github-copilot-token.js", () => ({
+          DEFAULT_COPILOT_API_BASE_URL: "https://api.individual.githubcopilot.com",
+          resolveCopilotApiToken: vi.fn().mockResolvedValue({
+            token: "copilot",
+            expiresAt: Date.now() + 60 * 60 * 1000,
+            source: "mock",
+            baseUrl: "https://api.copilot.example",
+          }),
+        }));
+
        const { ensureClawdbotModelsJson } = await import("./models-config.js");

        const agentDir = path.join(home, "agent-default-base-url");
@@ -62,55 +71,48 @@ describe("models-config", () => {
          providers: Record<string, { baseUrl?: string; models?: unknown[] }>;
        };

-        expect(parsed.providers["github-copilot"]?.baseUrl).toBe(DEFAULT_GITHUB_COPILOT_BASE_URL);
+        expect(parsed.providers["github-copilot"]?.baseUrl).toBe("https://api.copilot.example");
        expect(parsed.providers["github-copilot"]?.models?.length ?? 0).toBe(0);
      } finally {
        process.env.COPILOT_GITHUB_TOKEN = previous;
      }
    });
  });
-  it("uses enterprise URL from auth profiles to derive base URL", async () => {
+  it("prefers COPILOT_GITHUB_TOKEN over GH_TOKEN and GITHUB_TOKEN", async () => {
    await withTempHome(async () => {
+      const previous = process.env.COPILOT_GITHUB_TOKEN;
+      const previousGh = process.env.GH_TOKEN;
+      const previousGithub = process.env.GITHUB_TOKEN;
+      process.env.COPILOT_GITHUB_TOKEN = "copilot-token";
+      process.env.GH_TOKEN = "gh-token";
+      process.env.GITHUB_TOKEN = "github-token";
+
      try {
        vi.resetModules();

-        const agentDir = path.join(process.env.HOME ?? home, "agent-enterprise");
-        await fs.mkdir(agentDir, { recursive: true });
-        await fs.writeFile(
-          path.join(agentDir, "auth-profiles.json"),
-          JSON.stringify(
-            {
-              version: 1,
-              profiles: {
-                "github-copilot:github": {
-                  type: "oauth",
-                  provider: "github-copilot",
-                  refresh: "gh-token",
-                  access: "gh-token",
-                  expires: 0,
-                  enterpriseUrl: "company.ghe.com",
-                },
-              },
-            },
-            null,
-            2,
-          ),
-        );
+        const resolveCopilotApiToken = vi.fn().mockResolvedValue({
+          token: "copilot",
+          expiresAt: Date.now() + 60 * 60 * 1000,
+          source: "mock",
+          baseUrl: "https://api.copilot.example",
+        });
+
+        vi.doMock("../providers/github-copilot-token.js", () => ({
+          DEFAULT_COPILOT_API_BASE_URL: "https://api.individual.githubcopilot.com",
+          resolveCopilotApiToken,
+        }));

        const { ensureClawdbotModelsJson } = await import("./models-config.js");

-        await ensureClawdbotModelsJson({ models: { providers: {} } }, agentDir);
+        await ensureClawdbotModelsJson({ models: { providers: {} } });

-        const raw = await fs.readFile(path.join(agentDir, "models.json"), "utf8");
-        const parsed = JSON.parse(raw) as {
-          providers: Record<string, { baseUrl?: string }>;
-        };
-
-        expect(parsed.providers["github-copilot"]?.baseUrl).toBe(
-          "https://copilot-api.company.ghe.com",
+        expect(resolveCopilotApiToken).toHaveBeenCalledWith(
+          expect.objectContaining({ githubToken: "copilot-token" }),
        );
      } finally {
-        // no-op
+        process.env.COPILOT_GITHUB_TOKEN = previous;
+        process.env.GH_TOKEN = previousGh;
+        process.env.GITHUB_TOKEN = previousGithub;
      }
    });
  });
@@ -3,7 +3,6 @@ import path from "node:path";
 import { afterEach, beforeEach, describe, expect, it, vi } from "vitest";
 import { withTempHome as withTempHomeBase } from "../../test/helpers/temp-home.js";
 import type { ClawdbotConfig } from "../config/config.js";
-import { DEFAULT_GITHUB_COPILOT_BASE_URL } from "../providers/github-copilot-utils.js";

 async function withTempHome<T>(fn: (home: string) => Promise<T>): Promise<T> {
  return withTempHomeBase(fn, { prefix: "clawdbot-models-" });
@@ -44,7 +43,7 @@ describe("models-config", () => {
    process.env.HOME = previousHome;
  });

-  it("uses default baseUrl when env token is present", async () => {
+  it("falls back to default baseUrl when token exchange fails", async () => {
    await withTempHome(async () => {
      const previous = process.env.COPILOT_GITHUB_TOKEN;
      process.env.COPILOT_GITHUB_TOKEN = "gh-token";
@@ -52,6 +51,11 @@ describe("models-config", () => {
      try {
        vi.resetModules();

+        vi.doMock("../providers/github-copilot-token.js", () => ({
+          DEFAULT_COPILOT_API_BASE_URL: "https://api.default.test",
+          resolveCopilotApiToken: vi.fn().mockRejectedValue(new Error("boom")),
+        }));
+
        const { ensureClawdbotModelsJson } = await import("./models-config.js");
        const { resolveClawdbotAgentDir } = await import("./agent-paths.js");

@@ -63,13 +67,13 @@ describe("models-config", () => {
          providers: Record<string, { baseUrl?: string }>;
        };

-        expect(parsed.providers["github-copilot"]?.baseUrl).toBe(DEFAULT_GITHUB_COPILOT_BASE_URL);
+        expect(parsed.providers["github-copilot"]?.baseUrl).toBe("https://api.default.test");
      } finally {
        process.env.COPILOT_GITHUB_TOKEN = previous;
      }
    });
  });
-  it("normalizes enterprise URL when deriving base URL", async () => {
+  it("uses agentDir override auth profiles for copilot injection", async () => {
    await withTempHome(async (home) => {
      const previous = process.env.COPILOT_GITHUB_TOKEN;
      const previousGh = process.env.GH_TOKEN;
@@ -90,12 +94,9 @@ describe("models-config", () => {
              version: 1,
              profiles: {
                "github-copilot:github": {
-                  type: "oauth",
+                  type: "token",
                  provider: "github-copilot",
-                  refresh: "gh-profile-token",
-                  access: "gh-profile-token",
-                  expires: 0,
-                  enterpriseUrl: "https://company.ghe.com/",
+                  token: "gh-profile-token",
                },
              },
            },
@@ -104,6 +105,16 @@ describe("models-config", () => {
          ),
        );

+        vi.doMock("../providers/github-copilot-token.js", () => ({
+          DEFAULT_COPILOT_API_BASE_URL: "https://api.individual.githubcopilot.com",
+          resolveCopilotApiToken: vi.fn().mockResolvedValue({
+            token: "copilot",
+            expiresAt: Date.now() + 60 * 60 * 1000,
+            source: "mock",
+            baseUrl: "https://api.copilot.example",
+          }),
+        }));
+
        const { ensureClawdbotModelsJson } = await import("./models-config.js");

        await ensureClawdbotModelsJson({ models: { providers: {} } }, agentDir);
@@ -113,9 +124,7 @@ describe("models-config", () => {
          providers: Record<string, { baseUrl?: string }>;
        };

-        expect(parsed.providers["github-copilot"]?.baseUrl).toBe(
-          "https://copilot-api.company.ghe.com",
-        );
+        expect(parsed.providers["github-copilot"]?.baseUrl).toBe("https://api.copilot.example");
      } finally {
        if (previous === undefined) delete process.env.COPILOT_GITHUB_TOKEN;
        else process.env.COPILOT_GITHUB_TOKEN = previous;
@@ -1,8 +1,8 @@
 import type { ClawdbotConfig } from "../config/config.js";
 import {
-  normalizeGithubCopilotDomain,
-  resolveGithubCopilotBaseUrl,
-} from "../providers/github-copilot-utils.js";
+  DEFAULT_COPILOT_API_BASE_URL,
+  resolveCopilotApiToken,
+} from "../providers/github-copilot-token.js";
 import { ensureAuthProfileStore, listProfilesForProvider } from "./auth-profiles.js";
 import { resolveAwsSdkEnvVarName, resolveEnvApiKey } from "./model-auth.js";
 import {
@@ -331,18 +331,29 @@ export async function resolveImplicitCopilotProvider(params: {

  if (!hasProfile && !githubToken) return null;

-  let enterpriseDomain: string | null = null;
-  if (hasProfile) {
+  let selectedGithubToken = githubToken;
+  if (!selectedGithubToken && hasProfile) {
    // Use the first available profile as a default for discovery (it will be
    // re-resolved per-run by the embedded runner).
    const profileId = listProfilesForProvider(authStore, "github-copilot")[0];
    const profile = profileId ? authStore.profiles[profileId] : undefined;
-    if (profile && "enterpriseUrl" in profile && typeof profile.enterpriseUrl === "string") {
-      enterpriseDomain = normalizeGithubCopilotDomain(profile.enterpriseUrl);
+    if (profile && profile.type === "token") {
+      selectedGithubToken = profile.token;
    }
  }

-  const baseUrl = resolveGithubCopilotBaseUrl(enterpriseDomain);
+  let baseUrl = DEFAULT_COPILOT_API_BASE_URL;
+  if (selectedGithubToken) {
+    try {
+      const token = await resolveCopilotApiToken({
+        githubToken: selectedGithubToken,
+        env,
+      });
+      baseUrl = token.baseUrl;
+    } catch {
+      baseUrl = DEFAULT_COPILOT_API_BASE_URL;
+    }
+  }

  // pi-coding-agent's ModelRegistry marks a model "available" only if its
  // `AuthStorage` has auth configured for that provider (via auth.json/env/etc).
@@ -353,7 +364,7 @@ export async function resolveImplicitCopilotProvider(params: {
  // GitHub token (not the exchanged Copilot token), and (3) matches existing
  // patterns for OAuth-like providers in pi-coding-agent.
  // Note: we deliberately do not write pi-coding-agent's `auth.json` here.
-  // Clawdbot uses its own auth store and passes the GitHub token at runtime.
+  // Clawdbot uses its own auth store and exchanges tokens at runtime.
  // `models list` uses Clawdbot's auth heuristics for availability.

  // We intentionally do NOT define custom models for Copilot in models.json.
@@ -3,7 +3,6 @@ import path from "node:path";
 import { afterEach, beforeEach, describe, expect, it, vi } from "vitest";
 import { withTempHome as withTempHomeBase } from "../../test/helpers/temp-home.js";
 import type { ClawdbotConfig } from "../config/config.js";
-import { DEFAULT_GITHUB_COPILOT_BASE_URL } from "../providers/github-copilot-utils.js";

 async function withTempHome<T>(fn: (home: string) => Promise<T>): Promise<T> {
  return withTempHomeBase(fn, { prefix: "clawdbot-models-" });
@@ -81,16 +80,25 @@ describe("models-config", () => {
          ),
        );

+        const resolveCopilotApiToken = vi.fn().mockResolvedValue({
+          token: "copilot",
+          expiresAt: Date.now() + 60 * 60 * 1000,
+          source: "mock",
+          baseUrl: "https://api.copilot.example",
+        });
+
+        vi.doMock("../providers/github-copilot-token.js", () => ({
+          DEFAULT_COPILOT_API_BASE_URL: "https://api.individual.githubcopilot.com",
+          resolveCopilotApiToken,
+        }));
+
        const { ensureClawdbotModelsJson } = await import("./models-config.js");

        await ensureClawdbotModelsJson({ models: { providers: {} } }, agentDir);

-        const raw = await fs.readFile(path.join(agentDir, "models.json"), "utf8");
-        const parsed = JSON.parse(raw) as {
-          providers: Record<string, { baseUrl?: string }>;
-        };
-
-        expect(parsed.providers["github-copilot"]?.baseUrl).toBe(DEFAULT_GITHUB_COPILOT_BASE_URL);
+        expect(resolveCopilotApiToken).toHaveBeenCalledWith(
+          expect.objectContaining({ githubToken: "alpha-token" }),
+        );
      } finally {
        if (previous === undefined) delete process.env.COPILOT_GITHUB_TOKEN;
        else process.env.COPILOT_GITHUB_TOKEN = previous;
@@ -109,6 +117,16 @@ describe("models-config", () => {
      try {
        vi.resetModules();

+        vi.doMock("../providers/github-copilot-token.js", () => ({
+          DEFAULT_COPILOT_API_BASE_URL: "https://api.individual.githubcopilot.com",
+          resolveCopilotApiToken: vi.fn().mockResolvedValue({
+            token: "copilot",
+            expiresAt: Date.now() + 60 * 60 * 1000,
+            source: "mock",
+            baseUrl: "https://api.copilot.example",
+          }),
+        }));
+
        const { ensureClawdbotModelsJson } = await import("./models-config.js");
        const { resolveClawdbotAgentDir } = await import("./agent-paths.js");

@@ -41,12 +41,11 @@ describe("resolveOpencodeZenAlias", () => {
 describe("resolveOpencodeZenModelApi", () => {
  it("maps APIs by model family", () => {
    expect(resolveOpencodeZenModelApi("claude-opus-4-5")).toBe("anthropic-messages");
-    expect(resolveOpencodeZenModelApi("minimax-m2.1-free")).toBe("anthropic-messages");
    expect(resolveOpencodeZenModelApi("gemini-3-pro")).toBe("google-generative-ai");
    expect(resolveOpencodeZenModelApi("gpt-5.2")).toBe("openai-responses");
    expect(resolveOpencodeZenModelApi("alpha-gd4")).toBe("openai-completions");
    expect(resolveOpencodeZenModelApi("big-pickle")).toBe("openai-completions");
-    expect(resolveOpencodeZenModelApi("glm-4.7-free")).toBe("openai-completions");
+    expect(resolveOpencodeZenModelApi("glm-4.7")).toBe("openai-completions");
    expect(resolveOpencodeZenModelApi("some-unknown-model")).toBe("openai-completions");
  });
 });
@@ -55,10 +54,10 @@ describe("getOpencodeZenStaticFallbackModels", () => {
  it("returns an array of models", () => {
    const models = getOpencodeZenStaticFallbackModels();
    expect(Array.isArray(models)).toBe(true);
-    expect(models.length).toBe(11);
+    expect(models.length).toBe(10);
  });

-  it("includes Claude, GPT, Gemini, GLM, and MiniMax models", () => {
+  it("includes Claude, GPT, Gemini, and GLM models", () => {
    const models = getOpencodeZenStaticFallbackModels();
    const ids = models.map((m) => m.id);

@@ -66,8 +65,7 @@ describe("getOpencodeZenStaticFallbackModels", () => {
    expect(ids).toContain("gpt-5.2");
    expect(ids).toContain("gpt-5.1-codex");
    expect(ids).toContain("gemini-3-pro");
-    expect(ids).toContain("glm-4.7-free");
-    expect(ids).toContain("minimax-m2.1-free");
+    expect(ids).toContain("glm-4.7");
  });

  it("returns valid ModelDefinitionConfig objects", () => {
@@ -90,8 +88,7 @@ describe("OPENCODE_ZEN_MODEL_ALIASES", () => {
    expect(OPENCODE_ZEN_MODEL_ALIASES.codex).toBe("gpt-5.1-codex");
    expect(OPENCODE_ZEN_MODEL_ALIASES.gpt5).toBe("gpt-5.2");
    expect(OPENCODE_ZEN_MODEL_ALIASES.gemini).toBe("gemini-3-pro");
-    expect(OPENCODE_ZEN_MODEL_ALIASES.glm).toBe("glm-4.7-free");
-    expect(OPENCODE_ZEN_MODEL_ALIASES.minimax).toBe("minimax-m2.1-free");
+    expect(OPENCODE_ZEN_MODEL_ALIASES.glm).toBe("glm-4.7");

    // Legacy aliases (kept for backward compatibility).
    expect(OPENCODE_ZEN_MODEL_ALIASES.sonnet).toBe("claude-opus-4-5");
@@ -68,13 +68,9 @@ export const OPENCODE_ZEN_MODEL_ALIASES: Record<string, string> = {
  "gemini-2.5-flash": "gemini-3-flash",

  // GLM (free + alpha)
-  glm: "glm-4.7-free",
-  "glm-free": "glm-4.7-free",
+  glm: "glm-4.7",
+  "glm-free": "glm-4.7",
  "alpha-glm": "alpha-glm-4.7",
-
-  // MiniMax
-  minimax: "minimax-m2.1-free",
-  "minimax-free": "minimax-m2.1-free",
 };

 /**
@@ -134,7 +130,7 @@ const MODEL_COSTS: Record<
    cacheWrite: 0,
  },
  "gpt-5.1": { input: 1.07, output: 8.5, cacheRead: 0.107, cacheWrite: 0 },
-  "glm-4.7-free": { input: 0, output: 0, cacheRead: 0, cacheWrite: 0 },
+  "glm-4.7": { input: 0, output: 0, cacheRead: 0, cacheWrite: 0 },
  "gemini-3-flash": { input: 0.5, output: 3, cacheRead: 0.05, cacheWrite: 0 },
  "gpt-5.1-codex-max": {
    input: 1.25,
@@ -142,7 +138,6 @@ const MODEL_COSTS: Record<
    cacheRead: 0.125,
    cacheWrite: 0,
  },
-  "minimax-m2.1-free": { input: 0, output: 0, cacheRead: 0, cacheWrite: 0 },
  "gpt-5.2": { input: 1.75, output: 14, cacheRead: 0.175, cacheWrite: 0 },
 };

@@ -155,10 +150,9 @@ const MODEL_CONTEXT_WINDOWS: Record<string, number> = {
  "alpha-glm-4.7": 204800,
  "gpt-5.1-codex-mini": 400000,
  "gpt-5.1": 400000,
-  "glm-4.7-free": 204800,
+  "glm-4.7": 204800,
  "gemini-3-flash": 1048576,
  "gpt-5.1-codex-max": 400000,
-  "minimax-m2.1-free": 204800,
  "gpt-5.2": 400000,
 };

@@ -173,10 +167,9 @@ const MODEL_MAX_TOKENS: Record<string, number> = {
  "alpha-glm-4.7": 131072,
  "gpt-5.1-codex-mini": 128000,
  "gpt-5.1": 128000,
-  "glm-4.7-free": 131072,
+  "glm-4.7": 131072,
  "gemini-3-flash": 65536,
  "gpt-5.1-codex-max": 128000,
-  "minimax-m2.1-free": 131072,
  "gpt-5.2": 128000,
 };

@@ -211,10 +204,9 @@ const MODEL_NAMES: Record<string, string> = {
  "alpha-glm-4.7": "Alpha GLM-4.7",
  "gpt-5.1-codex-mini": "GPT-5.1 Codex Mini",
  "gpt-5.1": "GPT-5.1",
-  "glm-4.7-free": "GLM-4.7",
+  "glm-4.7": "GLM-4.7",
  "gemini-3-flash": "Gemini 3 Flash",
  "gpt-5.1-codex-max": "GPT-5.1 Codex Max",
-  "minimax-m2.1-free": "MiniMax M2.1",
  "gpt-5.2": "GPT-5.2",
 };

@@ -240,10 +232,9 @@ export function getOpencodeZenStaticFallbackModels(): ModelDefinitionConfig[] {
    "alpha-glm-4.7",
    "gpt-5.1-codex-mini",
    "gpt-5.1",
-    "glm-4.7-free",
+    "glm-4.7",
    "gemini-3-flash",
    "gpt-5.1-codex-max",
-    "minimax-m2.1-free",
    "gpt-5.2",
  ];

@@ -3,7 +3,7 @@ import os from "node:os";
 import path from "node:path";

 import type { AssistantMessage } from "@mariozechner/pi-ai";
-import { beforeEach, describe, expect, it, vi } from "vitest";
+import { beforeAll, beforeEach, describe, expect, it, vi } from "vitest";

 import type { ClawdbotConfig } from "../config/config.js";
 import type { EmbeddedRunAttemptResult } from "./pi-embedded-runner/run/types.js";
@@ -16,13 +16,15 @@ vi.mock("./pi-embedded-runner/run/attempt.js", () => ({

 let runEmbeddedPiAgent: typeof import("./pi-embedded-runner.js").runEmbeddedPiAgent;

-beforeEach(async () => {
-  vi.useRealTimers();
-  vi.resetModules();
-  runEmbeddedAttemptMock.mockReset();
+beforeAll(async () => {
  ({ runEmbeddedPiAgent } = await import("./pi-embedded-runner.js"));
 });

+beforeEach(() => {
+  vi.useRealTimers();
+  runEmbeddedAttemptMock.mockReset();
+});
+
 const baseUsage = {
  input: 0,
  output: 0,
@@ -128,6 +128,13 @@ export async function compactEmbeddedPiSession(params: {
              `No API key resolved for provider "${model.provider}" (auth mode: ${apiKeyInfo.mode}).`,
            );
          }
+        } else if (model.provider === "github-copilot") {
+          const { resolveCopilotApiToken } =
+            await import("../../providers/github-copilot-token.js");
+          const copilotToken = await resolveCopilotApiToken({
+            githubToken: apiKeyInfo.apiKey,
+          });
+          authStorage.setRuntimeApiKey(model.provider, copilotToken.token);
        } else {
          authStorage.setRuntimeApiKey(model.provider, apiKeyInfo.apiKey);
        }
@@ -1,3 +1,5 @@
+import { EventEmitter } from "node:events";
+
 import type { AgentMessage, AgentTool } from "@mariozechner/pi-agent-core";
 import type { TSchema } from "@sinclair/typebox";
 import type { SessionManager } from "@mariozechner/pi-coding-agent";
@@ -184,10 +186,28 @@ export function logToolSchemasForGoogle(params: { tools: AgentTool[]; provider:
  }
 }

+// Event emitter for unhandled compaction failures that escape try-catch blocks.
+// Listeners can use this to trigger session recovery with retry.
+const compactionFailureEmitter = new EventEmitter();
+
+export type CompactionFailureListener = (reason: string) => void;
+
+/**
+ * Register a listener for unhandled compaction failures.
+ * Called when auto-compaction fails in a way that escapes the normal try-catch,
+ * e.g., when the summarization request itself exceeds the model's token limit.
+ * Returns an unsubscribe function.
+ */
+export function onUnhandledCompactionFailure(cb: CompactionFailureListener): () => void {
+  compactionFailureEmitter.on("failure", cb);
+  return () => compactionFailureEmitter.off("failure", cb);
+}
+
 registerUnhandledRejectionHandler((reason) => {
  const message = describeUnknownError(reason);
  if (!isCompactionFailureError(message)) return false;
  log.error(`Auto-compaction failed (unhandled): ${message}`);
+  compactionFailureEmitter.emit("failure", message);
  return true;
 });

@@ -7,20 +7,9 @@ import { resolveClawdbotAgentDir } from "../agent-paths.js";
 import { DEFAULT_CONTEXT_TOKENS } from "../defaults.js";
 import { normalizeModelCompat } from "../model-compat.js";
 import { normalizeProviderId } from "../model-selection.js";
-import { resolveGithubCopilotUserAgent } from "../../providers/github-copilot-utils.js";

 type InlineModelEntry = ModelDefinitionConfig & { provider: string };

-function applyProviderModelOverrides(model: Model<Api>): Model<Api> {
-  if (model.provider === "github-copilot") {
-    const headers = model.headers
-      ? { ...model.headers, "User-Agent": resolveGithubCopilotUserAgent() }
-      : { "User-Agent": resolveGithubCopilotUserAgent() };
-    return { ...model, headers };
-  }
-  return model;
-}
-
 export function buildInlineProviderModels(
  providers: Record<string, { models?: ModelDefinitionConfig[] }>,
 ): InlineModelEntry[] {
@@ -71,7 +60,7 @@ export function resolveModel(
    if (inlineMatch) {
      const normalized = normalizeModelCompat(inlineMatch as Model<Api>);
      return {
-        model: applyProviderModelOverrides(normalized),
+        model: normalized,
        authStorage,
        modelRegistry,
      };
@@ -89,7 +78,7 @@ export function resolveModel(
        contextWindow: providerCfg?.models?.[0]?.contextWindow ?? DEFAULT_CONTEXT_TOKENS,
        maxTokens: providerCfg?.models?.[0]?.maxTokens ?? DEFAULT_CONTEXT_TOKENS,
      } as Model<Api>);
-      return { model: applyProviderModelOverrides(fallbackModel), authStorage, modelRegistry };
+      return { model: fallbackModel, authStorage, modelRegistry };
    }
    return {
      error: `Unknown model: ${provider}/${modelId}`,
@@ -97,9 +86,5 @@ export function resolveModel(
      modelRegistry,
    };
  }
-  return {
-    model: applyProviderModelOverrides(normalizeModelCompat(model)),
-    authStorage,
-    modelRegistry,
-  };
+  return { model: normalizeModelCompat(model), authStorage, modelRegistry };
 }
@@ -184,8 +184,17 @@ export async function runEmbeddedPiAgent(
          lastProfileId = resolvedProfileId;
          return;
        }
-        authStorage.setRuntimeApiKey(model.provider, apiKeyInfo.apiKey);
-        lastProfileId = resolvedProfileId;
+        if (model.provider === "github-copilot") {
+          const { resolveCopilotApiToken } =
+            await import("../../providers/github-copilot-token.js");
+          const copilotToken = await resolveCopilotApiToken({
+            githubToken: apiKeyInfo.apiKey,
+          });
+          authStorage.setRuntimeApiKey(model.provider, copilotToken.token);
+        } else {
+          authStorage.setRuntimeApiKey(model.provider, apiKeyInfo.apiKey);
+        }
+        lastProfileId = apiKeyInfo.profileId;
      };

      const advanceAuthProfile = async (): Promise<boolean> => {
@@ -109,14 +109,18 @@ export function setActiveEmbeddedRun(sessionId: string, handle: EmbeddedPiQueueH
    state: "processing",
    reason: wasActive ? "run_replaced" : "run_started",
  });
-  diag.info(`run registered: sessionId=${sessionId} totalActive=${ACTIVE_EMBEDDED_RUNS.size}`);
+  if (!sessionId.startsWith("probe-")) {
+    diag.info(`run registered: sessionId=${sessionId} totalActive=${ACTIVE_EMBEDDED_RUNS.size}`);
+  }
 }

 export function clearActiveEmbeddedRun(sessionId: string, handle: EmbeddedPiQueueHandle) {
  if (ACTIVE_EMBEDDED_RUNS.get(sessionId) === handle) {
    ACTIVE_EMBEDDED_RUNS.delete(sessionId);
    logSessionStateChange({ sessionId, state: "idle", reason: "run_completed" });
-    diag.info(`run cleared: sessionId=${sessionId} totalActive=${ACTIVE_EMBEDDED_RUNS.size}`);
+    if (!sessionId.startsWith("probe-")) {
+      diag.info(`run cleared: sessionId=${sessionId} totalActive=${ACTIVE_EMBEDDED_RUNS.size}`);
+    }
    notifyEmbeddedRunEnded(sessionId);
  } else {
    diag.debug(`run clear skipped: sessionId=${sessionId} reason=handle_mismatch`);
@@ -9,7 +9,7 @@ import { formatToolDetail, resolveToolDisplay } from "./tool-display.js";
 * - <invoke name="...">...</invoke> blocks
 * - </minimax:tool_call> closing tags
 */
-function stripMinimaxToolCallXml(text: string): string {
+export function stripMinimaxToolCallXml(text: string): string {
  if (!text) return text;
  if (!/minimax:tool_call/i.test(text)) return text;

@@ -28,7 +28,7 @@ function stripMinimaxToolCallXml(text: string): string {
 * downgraded to text blocks like `[Tool Call: name (ID: ...)]`. These should
 * not be shown to users.
 */
-function stripDowngradedToolCallText(text: string): string {
+export function stripDowngradedToolCallText(text: string): string {
  if (!text) return text;
  if (!/\[Tool (?:Call|Result)/i.test(text)) return text;

@@ -165,7 +165,7 @@ function stripDowngradedToolCallText(text: string): string {
 * This is a safety net for cases where the model outputs <think> tags
 * that slip through other filtering mechanisms.
 */
-function stripThinkingTagsFromText(text: string): string {
+export function stripThinkingTagsFromText(text: string): string {
  if (!text) return text;
  // Quick check to avoid regex overhead when no tags present.
  if (!/(?:think(?:ing)?|thought|antthinking)/i.test(text)) return text;
@@ -3,7 +3,15 @@ import { describe, expect, it } from "vitest";

 import { __testing } from "./compaction-safeguard.js";

-const { collectToolFailures, formatToolFailuresSection } = __testing;
+const {
+  collectToolFailures,
+  formatToolFailuresSection,
+  computeAdaptiveChunkRatio,
+  isOversizedForSummary,
+  BASE_CHUNK_RATIO,
+  MIN_CHUNK_RATIO,
+  SAFETY_MARGIN,
+} = __testing;

 describe("compaction-safeguard tool failures", () => {
  it("formats tool failures with meta and summary", () => {
@@ -96,3 +104,107 @@ describe("compaction-safeguard tool failures", () => {
    expect(section).toBe("");
  });
 });
+
+describe("computeAdaptiveChunkRatio", () => {
+  const CONTEXT_WINDOW = 200_000;
+
+  it("returns BASE_CHUNK_RATIO for normal messages", () => {
+    // Small messages: 1000 tokens each, well under 10% of context
+    const messages: AgentMessage[] = [
+      { role: "user", content: "x".repeat(1000), timestamp: Date.now() },
+      {
+        role: "assistant",
+        content: [{ type: "text", text: "y".repeat(1000) }],
+        timestamp: Date.now(),
+      },
+    ];
+
+    const ratio = computeAdaptiveChunkRatio(messages, CONTEXT_WINDOW);
+    expect(ratio).toBe(BASE_CHUNK_RATIO);
+  });
+
+  it("reduces ratio when average message > 10% of context", () => {
+    // Large messages: ~50K tokens each (25% of context)
+    const messages: AgentMessage[] = [
+      { role: "user", content: "x".repeat(50_000 * 4), timestamp: Date.now() },
+      {
+        role: "assistant",
+        content: [{ type: "text", text: "y".repeat(50_000 * 4) }],
+        timestamp: Date.now(),
+      },
+    ];
+
+    const ratio = computeAdaptiveChunkRatio(messages, CONTEXT_WINDOW);
+    expect(ratio).toBeLessThan(BASE_CHUNK_RATIO);
+    expect(ratio).toBeGreaterThanOrEqual(MIN_CHUNK_RATIO);
+  });
+
+  it("respects MIN_CHUNK_RATIO floor", () => {
+    // Very large messages that would push ratio below minimum
+    const messages: AgentMessage[] = [
+      { role: "user", content: "x".repeat(150_000 * 4), timestamp: Date.now() },
+    ];
+
+    const ratio = computeAdaptiveChunkRatio(messages, CONTEXT_WINDOW);
+    expect(ratio).toBeGreaterThanOrEqual(MIN_CHUNK_RATIO);
+  });
+
+  it("handles empty message array", () => {
+    const ratio = computeAdaptiveChunkRatio([], CONTEXT_WINDOW);
+    expect(ratio).toBe(BASE_CHUNK_RATIO);
+  });
+
+  it("handles single huge message", () => {
+    // Single massive message
+    const messages: AgentMessage[] = [
+      { role: "user", content: "x".repeat(180_000 * 4), timestamp: Date.now() },
+    ];
+
+    const ratio = computeAdaptiveChunkRatio(messages, CONTEXT_WINDOW);
+    expect(ratio).toBeGreaterThanOrEqual(MIN_CHUNK_RATIO);
+    expect(ratio).toBeLessThanOrEqual(BASE_CHUNK_RATIO);
+  });
+});
+
+describe("isOversizedForSummary", () => {
+  const CONTEXT_WINDOW = 200_000;
+
+  it("returns false for small messages", () => {
+    const msg: AgentMessage = {
+      role: "user",
+      content: "Hello, world!",
+      timestamp: Date.now(),
+    };
+
+    expect(isOversizedForSummary(msg, CONTEXT_WINDOW)).toBe(false);
+  });
+
+  it("returns true for messages > 50% of context", () => {
+    // Message with ~120K tokens (60% of 200K context)
+    // After safety margin (1.2x), effective is 144K which is > 100K (50%)
+    const msg: AgentMessage = {
+      role: "user",
+      content: "x".repeat(120_000 * 4),
+      timestamp: Date.now(),
+    };
+
+    expect(isOversizedForSummary(msg, CONTEXT_WINDOW)).toBe(true);
+  });
+
+  it("applies safety margin", () => {
+    // Message at exactly 50% of context before margin
+    // After SAFETY_MARGIN (1.2), it becomes 60% which is > 50%
+    const halfContextChars = (CONTEXT_WINDOW * 0.5) / SAFETY_MARGIN;
+    const msg: AgentMessage = {
+      role: "user",
+      content: "x".repeat(Math.floor(halfContextChars * 4)),
+      timestamp: Date.now(),
+    };
+
+    // With safety margin applied, this should be at the boundary
+    // The function checks if tokens * SAFETY_MARGIN > contextWindow * 0.5
+    const isOversized = isOversizedForSummary(msg, CONTEXT_WINDOW);
+    // Due to token estimation, this could be either true or false at the boundary
+    expect(typeof isOversized).toBe("boolean");
+  });
+});
@@ -4,7 +4,9 @@ import { estimateTokens, generateSummary } from "@mariozechner/pi-coding-agent";

 import { DEFAULT_CONTEXT_TOKENS } from "../defaults.js";

-const MAX_CHUNK_RATIO = 0.4;
+const BASE_CHUNK_RATIO = 0.4;
+const MIN_CHUNK_RATIO = 0.15;
+const SAFETY_MARGIN = 1.2; // 20% buffer for estimateTokens() inaccuracy
 const FALLBACK_SUMMARY =
  "Summary unavailable due to context limits. Older messages were truncated.";
 const TURN_PREFIX_INSTRUCTIONS =
@@ -160,6 +162,38 @@ function chunkMessages(messages: AgentMessage[], maxTokens: number): AgentMessag
  return chunks;
 }

+/**
+ * Compute adaptive chunk ratio based on average message size.
+ * When messages are large, we use smaller chunks to avoid exceeding model limits.
+ */
+function computeAdaptiveChunkRatio(messages: AgentMessage[], contextWindow: number): number {
+  if (messages.length === 0) return BASE_CHUNK_RATIO;
+
+  const totalTokens = messages.reduce((sum, m) => sum + estimateTokens(m), 0);
+  const avgTokens = totalTokens / messages.length;
+
+  // Apply safety margin to account for estimation inaccuracy
+  const safeAvgTokens = avgTokens * SAFETY_MARGIN;
+  const avgRatio = safeAvgTokens / contextWindow;
+
+  // If average message is > 10% of context, reduce chunk ratio
+  if (avgRatio > 0.1) {
+    const reduction = Math.min(avgRatio * 2, BASE_CHUNK_RATIO - MIN_CHUNK_RATIO);
+    return Math.max(MIN_CHUNK_RATIO, BASE_CHUNK_RATIO - reduction);
+  }
+
+  return BASE_CHUNK_RATIO;
+}
+
+/**
+ * Check if a single message is too large to summarize.
+ * If single message > 50% of context, it can't be summarized safely.
+ */
+function isOversizedForSummary(msg: AgentMessage, contextWindow: number): boolean {
+  const tokens = estimateTokens(msg) * SAFETY_MARGIN;
+  return tokens > contextWindow * 0.5;
+}
+
 async function summarizeChunks(params: {
  messages: AgentMessage[];
  model: NonNullable<ExtensionContext["model"]>;
@@ -192,6 +226,78 @@ async function summarizeChunks(params: {
  return summary ?? "No prior history.";
 }

+/**
+ * Summarize with progressive fallback for handling oversized messages.
+ * If full summarization fails, tries partial summarization excluding oversized messages.
+ */
+async function summarizeWithFallback(params: {
+  messages: AgentMessage[];
+  model: NonNullable<ExtensionContext["model"]>;
+  apiKey: string;
+  signal: AbortSignal;
+  reserveTokens: number;
+  maxChunkTokens: number;
+  contextWindow: number;
+  customInstructions?: string;
+  previousSummary?: string;
+}): Promise<string> {
+  const { messages, contextWindow } = params;
+
+  if (messages.length === 0) {
+    return params.previousSummary ?? "No prior history.";
+  }
+
+  // Try full summarization first
+  try {
+    return await summarizeChunks(params);
+  } catch (fullError) {
+    console.warn(
+      `Full summarization failed, trying partial: ${
+        fullError instanceof Error ? fullError.message : String(fullError)
+      }`,
+    );
+  }
+
+  // Fallback 1: Summarize only small messages, note oversized ones
+  const smallMessages: AgentMessage[] = [];
+  const oversizedNotes: string[] = [];
+
+  for (const msg of messages) {
+    if (isOversizedForSummary(msg, contextWindow)) {
+      const role = (msg as { role?: string }).role ?? "message";
+      const tokens = estimateTokens(msg);
+      oversizedNotes.push(
+        `[Large ${role} (~${Math.round(tokens / 1000)}K tokens) omitted from summary]`,
+      );
+    } else {
+      smallMessages.push(msg);
+    }
+  }
+
+  if (smallMessages.length > 0) {
+    try {
+      const partialSummary = await summarizeChunks({
+        ...params,
+        messages: smallMessages,
+      });
+      const notes = oversizedNotes.length > 0 ? `\n\n${oversizedNotes.join("\n")}` : "";
+      return partialSummary + notes;
+    } catch (partialError) {
+      console.warn(
+        `Partial summarization also failed: ${
+          partialError instanceof Error ? partialError.message : String(partialError)
+        }`,
+      );
+    }
+  }
+
+  // Final fallback: Just note what was there
+  return (
+    `Context contained ${messages.length} messages (${oversizedNotes.length} oversized). ` +
+    `Summary unavailable due to size limits.`
+  );
+}
+
 export default function compactionSafeguardExtension(api: ExtensionAPI): void {
  api.on("session_before_compact", async (event, ctx) => {
    const { preparation, customInstructions, signal } = event;
@@ -233,29 +339,35 @@ export default function compactionSafeguardExtension(api: ExtensionAPI): void {
        1,
        Math.floor(model.contextWindow ?? DEFAULT_CONTEXT_TOKENS),
      );
-      const maxChunkTokens = Math.max(1, Math.floor(contextWindowTokens * MAX_CHUNK_RATIO));
+
+      // Use adaptive chunk ratio based on message sizes
+      const allMessages = [...preparation.messagesToSummarize, ...preparation.turnPrefixMessages];
+      const adaptiveRatio = computeAdaptiveChunkRatio(allMessages, contextWindowTokens);
+      const maxChunkTokens = Math.max(1, Math.floor(contextWindowTokens * adaptiveRatio));
      const reserveTokens = Math.max(1, Math.floor(preparation.settings.reserveTokens));

-      const historySummary = await summarizeChunks({
+      const historySummary = await summarizeWithFallback({
        messages: preparation.messagesToSummarize,
        model,
        apiKey,
        signal,
        reserveTokens,
        maxChunkTokens,
+        contextWindow: contextWindowTokens,
        customInstructions,
        previousSummary: preparation.previousSummary,
      });

      let summary = historySummary;
      if (preparation.isSplitTurn && preparation.turnPrefixMessages.length > 0) {
-        const prefixSummary = await summarizeChunks({
+        const prefixSummary = await summarizeWithFallback({
          messages: preparation.turnPrefixMessages,
          model,
          apiKey,
          signal,
          reserveTokens,
          maxChunkTokens,
+          contextWindow: contextWindowTokens,
          customInstructions: TURN_PREFIX_INSTRUCTIONS,
        });
        summary = `${historySummary}\n\n---\n\n**Turn Context (split turn):**\n\n${prefixSummary}`;
@@ -293,4 +405,9 @@ export default function compactionSafeguardExtension(api: ExtensionAPI): void {
 export const __testing = {
  collectToolFailures,
  formatToolFailuresSection,
+  computeAdaptiveChunkRatio,
+  isOversizedForSummary,
+  BASE_CHUNK_RATIO,
+  MIN_CHUNK_RATIO,
+  SAFETY_MARGIN,
 } as const;
@@ -1,212 +0,0 @@
-import type { AgentTool } from "@mariozechner/pi-agent-core";
-import { describe, expect, it, vi } from "vitest";
-import { __testing, createClawdbotCodingTools } from "./pi-tools.js";
-import { createBrowserTool } from "./tools/browser-tool.js";
-
-describe("createClawdbotCodingTools", () => {
-  describe("Claude/Gemini alias support", () => {
-    it("adds Claude-style aliases to schemas without dropping metadata", () => {
-      const base: AgentTool = {
-        name: "write",
-        description: "test",
-        parameters: {
-          type: "object",
-          required: ["path", "content"],
-          properties: {
-            path: { type: "string", description: "Path" },
-            content: { type: "string", description: "Body" },
-          },
-        },
-        execute: vi.fn(),
-      };
-
-      const patched = __testing.patchToolSchemaForClaudeCompatibility(base);
-      const params = patched.parameters as {
-        properties?: Record<string, unknown>;
-        required?: string[];
-      };
-      const props = params.properties ?? {};
-
-      expect(props.file_path).toEqual(props.path);
-      expect(params.required ?? []).not.toContain("path");
-      expect(params.required ?? []).not.toContain("file_path");
-    });
-
-    it("normalizes file_path to path and enforces required groups at runtime", async () => {
-      const execute = vi.fn(async (_id, args) => args);
-      const tool: AgentTool = {
-        name: "write",
-        description: "test",
-        parameters: {
-          type: "object",
-          required: ["path", "content"],
-          properties: {
-            path: { type: "string" },
-            content: { type: "string" },
-          },
-        },
-        execute,
-      };
-
-      const wrapped = __testing.wrapToolParamNormalization(tool, [{ keys: ["path", "file_path"] }]);
-
-      await wrapped.execute("tool-1", { file_path: "foo.txt", content: "x" });
-      expect(execute).toHaveBeenCalledWith(
-        "tool-1",
-        { path: "foo.txt", content: "x" },
-        undefined,
-        undefined,
-      );
-
-      await expect(wrapped.execute("tool-2", { content: "x" })).rejects.toThrow(
-        /Missing required parameter/,
-      );
-      await expect(wrapped.execute("tool-3", { file_path: "   ", content: "x" })).rejects.toThrow(
-        /Missing required parameter/,
-      );
-    });
-  });
-
-  it("keeps browser tool schema OpenAI-compatible without normalization", () => {
-    const browser = createBrowserTool();
-    const schema = browser.parameters as { type?: unknown; anyOf?: unknown };
-    expect(schema.type).toBe("object");
-    expect(schema.anyOf).toBeUndefined();
-  });
-  it("mentions Chrome extension relay in browser tool description", () => {
-    const browser = createBrowserTool();
-    expect(browser.description).toMatch(/Chrome extension/i);
-    expect(browser.description).toMatch(/profile="chrome"/i);
-  });
-  it("keeps browser tool schema properties after normalization", () => {
-    const tools = createClawdbotCodingTools();
-    const browser = tools.find((tool) => tool.name === "browser");
-    expect(browser).toBeDefined();
-    const parameters = browser?.parameters as {
-      anyOf?: unknown[];
-      properties?: Record<string, unknown>;
-      required?: string[];
-    };
-    expect(parameters.properties?.action).toBeDefined();
-    expect(parameters.properties?.target).toBeDefined();
-    expect(parameters.properties?.controlUrl).toBeDefined();
-    expect(parameters.properties?.targetUrl).toBeDefined();
-    expect(parameters.properties?.request).toBeDefined();
-    expect(parameters.required ?? []).toContain("action");
-  });
-  it("exposes raw for gateway config.apply tool calls", () => {
-    const tools = createClawdbotCodingTools();
-    const gateway = tools.find((tool) => tool.name === "gateway");
-    expect(gateway).toBeDefined();
-
-    const parameters = gateway?.parameters as {
-      type?: unknown;
-      required?: string[];
-      properties?: Record<string, unknown>;
-    };
-    expect(parameters.type).toBe("object");
-    expect(parameters.properties?.raw).toBeDefined();
-    expect(parameters.required ?? []).not.toContain("raw");
-  });
-  it("flattens anyOf-of-literals to enum for provider compatibility", () => {
-    const tools = createClawdbotCodingTools();
-    const browser = tools.find((tool) => tool.name === "browser");
-    expect(browser).toBeDefined();
-
-    const parameters = browser?.parameters as {
-      properties?: Record<string, unknown>;
-    };
-    const action = parameters.properties?.action as
-      | {
-          type?: unknown;
-          enum?: unknown[];
-          anyOf?: unknown[];
-        }
-      | undefined;
-
-    expect(action?.type).toBe("string");
-    expect(action?.anyOf).toBeUndefined();
-    expect(Array.isArray(action?.enum)).toBe(true);
-    expect(action?.enum).toContain("act");
-
-    const snapshotFormat = parameters.properties?.snapshotFormat as
-      | {
-          type?: unknown;
-          enum?: unknown[];
-          anyOf?: unknown[];
-        }
-      | undefined;
-    expect(snapshotFormat?.type).toBe("string");
-    expect(snapshotFormat?.anyOf).toBeUndefined();
-    expect(snapshotFormat?.enum).toEqual(["aria", "ai"]);
-  });
-  it("inlines local $ref before removing unsupported keywords", () => {
-    const cleaned = __testing.cleanToolSchemaForGemini({
-      type: "object",
-      properties: {
-        foo: { $ref: "#/$defs/Foo" },
-      },
-      $defs: {
-        Foo: { type: "string", enum: ["a", "b"] },
-      },
-    }) as {
-      $defs?: unknown;
-      properties?: Record<string, unknown>;
-    };
-
-    expect(cleaned.$defs).toBeUndefined();
-    expect(cleaned.properties).toBeDefined();
-    expect(cleaned.properties?.foo).toMatchObject({
-      type: "string",
-      enum: ["a", "b"],
-    });
-  });
-  it("cleans tuple items schemas", () => {
-    const cleaned = __testing.cleanToolSchemaForGemini({
-      type: "object",
-      properties: {
-        tuples: {
-          type: "array",
-          items: [
-            { type: "string", format: "uuid" },
-            { type: "number", minimum: 1 },
-          ],
-        },
-      },
-    }) as {
-      properties?: Record<string, unknown>;
-    };
-
-    const tuples = cleaned.properties?.tuples as { items?: unknown } | undefined;
-    const items = Array.isArray(tuples?.items) ? tuples?.items : [];
-    const first = items[0] as { format?: unknown } | undefined;
-    const second = items[1] as { minimum?: unknown } | undefined;
-
-    expect(first?.format).toBeUndefined();
-    expect(second?.minimum).toBeUndefined();
-  });
-  it("drops null-only union variants without flattening other unions", () => {
-    const cleaned = __testing.cleanToolSchemaForGemini({
-      type: "object",
-      properties: {
-        parentId: { anyOf: [{ type: "string" }, { type: "null" }] },
-        count: { oneOf: [{ type: "string" }, { type: "number" }] },
-      },
-    }) as {
-      properties?: Record<string, unknown>;
-    };
-
-    const parentId = cleaned.properties?.parentId as
-      | { type?: unknown; anyOf?: unknown; oneOf?: unknown }
-      | undefined;
-    expect(parentId?.anyOf).toBeUndefined();
-    expect(parentId?.oneOf).toBeUndefined();
-    expect(parentId?.type).toBe("string");
-
-    const count = cleaned.properties?.count as
-      | { type?: unknown; anyOf?: unknown; oneOf?: unknown }
-      | undefined;
-    expect(count?.anyOf).toBeUndefined();
-    expect(Array.isArray(count?.oneOf)).toBe(true);
-  });
-});
@@ -1,74 +1,12 @@
-import type { AgentTool } from "@mariozechner/pi-agent-core";
-import { describe, expect, it, vi } from "vitest";
+import { describe, expect, it } from "vitest";
 import type { ClawdbotConfig } from "../config/config.js";
-import { __testing, createClawdbotCodingTools } from "./pi-tools.js";
+import "./test-helpers/fast-coding-tools.js";
+import { createClawdbotCodingTools } from "./pi-tools.js";
+
+const defaultTools = createClawdbotCodingTools();

 describe("createClawdbotCodingTools", () => {
-  describe("Claude/Gemini alias support", () => {
-    it("adds Claude-style aliases to schemas without dropping metadata", () => {
-      const base: AgentTool = {
-        name: "write",
-        description: "test",
-        parameters: {
-          type: "object",
-          required: ["path", "content"],
-          properties: {
-            path: { type: "string", description: "Path" },
-            content: { type: "string", description: "Body" },
-          },
-        },
-        execute: vi.fn(),
-      };
-
-      const patched = __testing.patchToolSchemaForClaudeCompatibility(base);
-      const params = patched.parameters as {
-        properties?: Record<string, unknown>;
-        required?: string[];
-      };
-      const props = params.properties ?? {};
-
-      expect(props.file_path).toEqual(props.path);
-      expect(params.required ?? []).not.toContain("path");
-      expect(params.required ?? []).not.toContain("file_path");
-    });
-
-    it("normalizes file_path to path and enforces required groups at runtime", async () => {
-      const execute = vi.fn(async (_id, args) => args);
-      const tool: AgentTool = {
-        name: "write",
-        description: "test",
-        parameters: {
-          type: "object",
-          required: ["path", "content"],
-          properties: {
-            path: { type: "string" },
-            content: { type: "string" },
-          },
-        },
-        execute,
-      };
-
-      const wrapped = __testing.wrapToolParamNormalization(tool, [{ keys: ["path", "file_path"] }]);
-
-      await wrapped.execute("tool-1", { file_path: "foo.txt", content: "x" });
-      expect(execute).toHaveBeenCalledWith(
-        "tool-1",
-        { path: "foo.txt", content: "x" },
-        undefined,
-        undefined,
-      );
-
-      await expect(wrapped.execute("tool-2", { content: "x" })).rejects.toThrow(
-        /Missing required parameter/,
-      );
-      await expect(wrapped.execute("tool-3", { file_path: "   ", content: "x" })).rejects.toThrow(
-        /Missing required parameter/,
-      );
-    });
-  });
-
  it("preserves action enums in normalized schemas", () => {
-    const tools = createClawdbotCodingTools();
    const toolNames = ["browser", "canvas", "nodes", "cron", "gateway", "message"];

    const collectActionValues = (schema: unknown, values: Set<string>): void => {
@@ -88,7 +26,7 @@ describe("createClawdbotCodingTools", () => {
    };

    for (const name of toolNames) {
-      const tool = tools.find((candidate) => candidate.name === name);
+      const tool = defaultTools.find((candidate) => candidate.name === name);
      expect(tool).toBeDefined();
      const parameters = tool?.parameters as {
        properties?: Record<string, unknown>;
@@ -108,10 +46,9 @@ describe("createClawdbotCodingTools", () => {
    }
  });
  it("includes exec and process tools by default", () => {
-    const tools = createClawdbotCodingTools();
-    expect(tools.some((tool) => tool.name === "exec")).toBe(true);
-    expect(tools.some((tool) => tool.name === "process")).toBe(true);
-    expect(tools.some((tool) => tool.name === "apply_patch")).toBe(false);
+    expect(defaultTools.some((tool) => tool.name === "exec")).toBe(true);
+    expect(defaultTools.some((tool) => tool.name === "process")).toBe(true);
+    expect(defaultTools.some((tool) => tool.name === "apply_patch")).toBe(false);
  });
  it("gates apply_patch behind tools.exec.applyPatch for OpenAI models", () => {
    const config: ClawdbotConfig = {
@@ -1,78 +1,16 @@
 import fs from "node:fs/promises";
 import os from "node:os";
 import path from "node:path";
-import type { AgentTool } from "@mariozechner/pi-agent-core";
 import sharp from "sharp";
-import { describe, expect, it, vi } from "vitest";
-import { __testing, createClawdbotCodingTools } from "./pi-tools.js";
+import { describe, expect, it } from "vitest";
+import "./test-helpers/fast-coding-tools.js";
+import { createClawdbotCodingTools } from "./pi-tools.js";
+
+const defaultTools = createClawdbotCodingTools();

 describe("createClawdbotCodingTools", () => {
-  describe("Claude/Gemini alias support", () => {
-    it("adds Claude-style aliases to schemas without dropping metadata", () => {
-      const base: AgentTool = {
-        name: "write",
-        description: "test",
-        parameters: {
-          type: "object",
-          required: ["path", "content"],
-          properties: {
-            path: { type: "string", description: "Path" },
-            content: { type: "string", description: "Body" },
-          },
-        },
-        execute: vi.fn(),
-      };
-
-      const patched = __testing.patchToolSchemaForClaudeCompatibility(base);
-      const params = patched.parameters as {
-        properties?: Record<string, unknown>;
-        required?: string[];
-      };
-      const props = params.properties ?? {};
-
-      expect(props.file_path).toEqual(props.path);
-      expect(params.required ?? []).not.toContain("path");
-      expect(params.required ?? []).not.toContain("file_path");
-    });
-
-    it("normalizes file_path to path and enforces required groups at runtime", async () => {
-      const execute = vi.fn(async (_id, args) => args);
-      const tool: AgentTool = {
-        name: "write",
-        description: "test",
-        parameters: {
-          type: "object",
-          required: ["path", "content"],
-          properties: {
-            path: { type: "string" },
-            content: { type: "string" },
-          },
-        },
-        execute,
-      };
-
-      const wrapped = __testing.wrapToolParamNormalization(tool, [{ keys: ["path", "file_path"] }]);
-
-      await wrapped.execute("tool-1", { file_path: "foo.txt", content: "x" });
-      expect(execute).toHaveBeenCalledWith(
-        "tool-1",
-        { path: "foo.txt", content: "x" },
-        undefined,
-        undefined,
-      );
-
-      await expect(wrapped.execute("tool-2", { content: "x" })).rejects.toThrow(
-        /Missing required parameter/,
-      );
-      await expect(wrapped.execute("tool-3", { file_path: "   ", content: "x" })).rejects.toThrow(
-        /Missing required parameter/,
-      );
-    });
-  });
-
  it("keeps read tool image metadata intact", async () => {
-    const tools = createClawdbotCodingTools();
-    const readTool = tools.find((tool) => tool.name === "read");
+    const readTool = defaultTools.find((tool) => tool.name === "read");
    expect(readTool).toBeDefined();

    const tmpDir = await fs.mkdtemp(path.join(os.tmpdir(), "clawdbot-read-"));
@@ -1,183 +0,0 @@
-import type { AgentTool } from "@mariozechner/pi-agent-core";
-import { describe, expect, it, vi } from "vitest";
-import { __testing, createClawdbotCodingTools } from "./pi-tools.js";
-
-describe("createClawdbotCodingTools", () => {
-  describe("Claude/Gemini alias support", () => {
-    it("adds Claude-style aliases to schemas without dropping metadata", () => {
-      const base: AgentTool = {
-        name: "write",
-        description: "test",
-        parameters: {
-          type: "object",
-          required: ["path", "content"],
-          properties: {
-            path: { type: "string", description: "Path" },
-            content: { type: "string", description: "Body" },
-          },
-        },
-        execute: vi.fn(),
-      };
-
-      const patched = __testing.patchToolSchemaForClaudeCompatibility(base);
-      const params = patched.parameters as {
-        properties?: Record<string, unknown>;
-        required?: string[];
-      };
-      const props = params.properties ?? {};
-
-      expect(props.file_path).toEqual(props.path);
-      expect(params.required ?? []).not.toContain("path");
-      expect(params.required ?? []).not.toContain("file_path");
-    });
-
-    it("normalizes file_path to path and enforces required groups at runtime", async () => {
-      const execute = vi.fn(async (_id, args) => args);
-      const tool: AgentTool = {
-        name: "write",
-        description: "test",
-        parameters: {
-          type: "object",
-          required: ["path", "content"],
-          properties: {
-            path: { type: "string" },
-            content: { type: "string" },
-          },
-        },
-        execute,
-      };
-
-      const wrapped = __testing.wrapToolParamNormalization(tool, [{ keys: ["path", "file_path"] }]);
-
-      await wrapped.execute("tool-1", { file_path: "foo.txt", content: "x" });
-      expect(execute).toHaveBeenCalledWith(
-        "tool-1",
-        { path: "foo.txt", content: "x" },
-        undefined,
-        undefined,
-      );
-
-      await expect(wrapped.execute("tool-2", { content: "x" })).rejects.toThrow(
-        /Missing required parameter/,
-      );
-      await expect(wrapped.execute("tool-3", { file_path: "   ", content: "x" })).rejects.toThrow(
-        /Missing required parameter/,
-      );
-    });
-  });
-
-  it("applies tool profiles before allow/deny policies", () => {
-    const tools = createClawdbotCodingTools({
-      config: { tools: { profile: "messaging" } },
-    });
-    const names = new Set(tools.map((tool) => tool.name));
-    expect(names.has("message")).toBe(true);
-    expect(names.has("sessions_send")).toBe(true);
-    expect(names.has("sessions_spawn")).toBe(false);
-    expect(names.has("exec")).toBe(false);
-    expect(names.has("browser")).toBe(false);
-  });
-  it("expands group shorthands in global tool policy", () => {
-    const tools = createClawdbotCodingTools({
-      config: { tools: { allow: ["group:fs"] } },
-    });
-    const names = new Set(tools.map((tool) => tool.name));
-    expect(names.has("read")).toBe(true);
-    expect(names.has("write")).toBe(true);
-    expect(names.has("edit")).toBe(true);
-    expect(names.has("exec")).toBe(false);
-    expect(names.has("browser")).toBe(false);
-  });
-  it("expands group shorthands in global tool deny policy", () => {
-    const tools = createClawdbotCodingTools({
-      config: { tools: { deny: ["group:fs"] } },
-    });
-    const names = new Set(tools.map((tool) => tool.name));
-    expect(names.has("read")).toBe(false);
-    expect(names.has("write")).toBe(false);
-    expect(names.has("edit")).toBe(false);
-    expect(names.has("exec")).toBe(true);
-  });
-  it("lets agent profiles override global profiles", () => {
-    const tools = createClawdbotCodingTools({
-      sessionKey: "agent:work:main",
-      config: {
-        tools: { profile: "coding" },
-        agents: {
-          list: [{ id: "work", tools: { profile: "messaging" } }],
-        },
-      },
-    });
-    const names = new Set(tools.map((tool) => tool.name));
-    expect(names.has("message")).toBe(true);
-    expect(names.has("exec")).toBe(false);
-    expect(names.has("read")).toBe(false);
-  });
-  it("removes unsupported JSON Schema keywords for Cloud Code Assist API compatibility", () => {
-    const tools = createClawdbotCodingTools();
-
-    // Helper to recursively check schema for unsupported keywords
-    const unsupportedKeywords = new Set([
-      "patternProperties",
-      "additionalProperties",
-      "$schema",
-      "$id",
-      "$ref",
-      "$defs",
-      "definitions",
-      "examples",
-      "minLength",
-      "maxLength",
-      "minimum",
-      "maximum",
-      "multipleOf",
-      "pattern",
-      "format",
-      "minItems",
-      "maxItems",
-      "uniqueItems",
-      "minProperties",
-      "maxProperties",
-    ]);
-
-    const findUnsupportedKeywords = (schema: unknown, path: string): string[] => {
-      const found: string[] = [];
-      if (!schema || typeof schema !== "object") return found;
-      if (Array.isArray(schema)) {
-        schema.forEach((item, i) => {
-          found.push(...findUnsupportedKeywords(item, `${path}[${i}]`));
-        });
-        return found;
-      }
-
-      const record = schema as Record<string, unknown>;
-      const properties =
-        record.properties &&
-        typeof record.properties === "object" &&
-        !Array.isArray(record.properties)
-          ? (record.properties as Record<string, unknown>)
-          : undefined;
-      if (properties) {
-        for (const [key, value] of Object.entries(properties)) {
-          found.push(...findUnsupportedKeywords(value, `${path}.properties.${key}`));
-        }
-      }
-
-      for (const [key, value] of Object.entries(record)) {
-        if (key === "properties") continue;
-        if (unsupportedKeywords.has(key)) {
-          found.push(`${path}.${key}`);
-        }
-        if (value && typeof value === "object") {
-          found.push(...findUnsupportedKeywords(value, `${path}.${key}`));
-        }
-      }
-      return found;
-    };
-
-    for (const tool of tools) {
-      const violations = findUnsupportedKeywords(tool.parameters, `${tool.name}.parameters`);
-      expect(violations).toEqual([]);
-    }
-  });
-});
@@ -1,74 +1,11 @@
 import fs from "node:fs/promises";
 import os from "node:os";
 import path from "node:path";
-import type { AgentTool } from "@mariozechner/pi-agent-core";
-import { describe, expect, it, vi } from "vitest";
-import { __testing, createClawdbotCodingTools } from "./pi-tools.js";
+import { describe, expect, it } from "vitest";
+import "./test-helpers/fast-coding-tools.js";
+import { createClawdbotCodingTools } from "./pi-tools.js";

 describe("createClawdbotCodingTools", () => {
-  describe("Claude/Gemini alias support", () => {
-    it("adds Claude-style aliases to schemas without dropping metadata", () => {
-      const base: AgentTool = {
-        name: "write",
-        description: "test",
-        parameters: {
-          type: "object",
-          required: ["path", "content"],
-          properties: {
-            path: { type: "string", description: "Path" },
-            content: { type: "string", description: "Body" },
-          },
-        },
-        execute: vi.fn(),
-      };
-
-      const patched = __testing.patchToolSchemaForClaudeCompatibility(base);
-      const params = patched.parameters as {
-        properties?: Record<string, unknown>;
-        required?: string[];
-      };
-      const props = params.properties ?? {};
-
-      expect(props.file_path).toEqual(props.path);
-      expect(params.required ?? []).not.toContain("path");
-      expect(params.required ?? []).not.toContain("file_path");
-    });
-
-    it("normalizes file_path to path and enforces required groups at runtime", async () => {
-      const execute = vi.fn(async (_id, args) => args);
-      const tool: AgentTool = {
-        name: "write",
-        description: "test",
-        parameters: {
-          type: "object",
-          required: ["path", "content"],
-          properties: {
-            path: { type: "string" },
-            content: { type: "string" },
-          },
-        },
-        execute,
-      };
-
-      const wrapped = __testing.wrapToolParamNormalization(tool, [{ keys: ["path", "file_path"] }]);
-
-      await wrapped.execute("tool-1", { file_path: "foo.txt", content: "x" });
-      expect(execute).toHaveBeenCalledWith(
-        "tool-1",
-        { path: "foo.txt", content: "x" },
-        undefined,
-        undefined,
-      );
-
-      await expect(wrapped.execute("tool-2", { content: "x" })).rejects.toThrow(
-        /Missing required parameter/,
-      );
-      await expect(wrapped.execute("tool-3", { file_path: "   ", content: "x" })).rejects.toThrow(
-        /Missing required parameter/,
-      );
-    });
-  });
-
  it("uses workspaceDir for Read tool path resolution", async () => {
    const tmpDir = await fs.mkdtemp(path.join(os.tmpdir(), "clawdbot-ws-"));
    try {
Author	SHA1	Message	Date
Peter Steinberger	0fbf728298	fix: sync macOS voice wake trigger words on blur/submit (#1506 ) (thanks @shiv19)	2026-01-23 19:54:01 +00:00
Shiva Prasad	16c4a3aea6	macOS: fix trigger word input disappearing when typing and on add Fixed issue where trigger words would disappear when typing or when adding new trigger words. The problem was that `swabbleTriggerWords` changes were triggering `VoiceWakeRuntime.refresh()` which sanitized the array by removing empty strings in real-time. Solution: Introduced local `@State` buffer `triggerEntries` with stable UUID identifiers for each trigger word entry. User edits now only affect the local state buffer and are synced back to `AppState` on explicit actions (submit, remove, disappear). This prevents premature sanitization during editing. The local state is loaded on view appear and when the view becomes active, ensuring it stays in sync with `AppState`.	2026-01-23 19:54:01 +00:00
Peter Steinberger	242add587f	fix: quiet auth probe diagnostics	2026-01-23 19:53:01 +00:00
Peter Steinberger	6fba598eaf	fix: handle gateway slash command replies in TUI	2026-01-23 19:48:22 +00:00
Peter Steinberger	75a54f0259	docs: note models usage suppression	2026-01-23 19:43:26 +00:00
Peter Steinberger	c63144ab14	fix: hide usage errors in status	2026-01-23 19:43:26 +00:00
Peter Steinberger	f07c39b265	docs: handle lint/format churn	2026-01-23 19:37:33 +00:00
Peter Steinberger	40181afded	feat: add models status auth probes	2026-01-23 19:28:55 +00:00
Peter Steinberger	2f1b9efe9a	style: wrap service path helpers	2026-01-23 19:17:57 +00:00
Peter Steinberger	ff30cef8a4	fix: expand linux service PATH handling	2026-01-23 19:16:41 +00:00
Robby	3d958d5466	fix(linux): add user bin directories to systemd service PATH for skill installation (#1512 ) * fix(linux): add user bin directories to systemd service PATH Fixes #1503 On Linux, the systemd service PATH was hardcoded to only include system directories (/usr/local/bin, /usr/bin, /bin), causing binaries installed via npm global with custom prefix or node version managers to not be found. This adds common Linux user bin directories to the PATH: - ~/.local/bin (XDG standard, pip, etc.) - ~/.npm-global/bin (npm custom prefix) - ~/bin (user's personal bin) - Node version manager paths (nvm, fnm, volta, asdf) - ~/.local/share/pnpm (pnpm global) - ~/.bun/bin (Bun) User directories are added before system directories so user-installed binaries take precedence. 🤖 AI-assisted (Claude Opus 4.5 via Clawdbot) 📋 Testing: Existing unit tests pass (7/7) * test: add comprehensive tests for Linux user bin directory resolution - Add dedicated tests for resolveLinuxUserBinDirs() function - Test path ordering (extraDirs > user dirs > system dirs) - Test buildMinimalServicePath() with HOME set/unset - Test platform-specific behavior (Linux vs macOS vs Windows) Test count: 7 → 20 (+13 tests) * test: add comprehensive tests for Linux user bin directory handling - Test Linux user directories included when HOME is set - Test Linux user directories excluded when HOME is missing - Test path ordering (extraDirs > user dirs > system dirs) - Test platform-specific behavior (Linux vs macOS vs Windows) - Test buildMinimalServicePath() with HOME in env Covers getMinimalServicePathParts() and buildMinimalServicePath() for all Linux user bin directory edge cases. Test count: 7 → 16 (+9 tests)	2026-01-23 19:06:14 +00:00
Peter Steinberger	cad7ed1cb8	fix(exec-approvals): stabilize allowlist ids (#1521 )	2026-01-23 19:00:45 +00:00
Peter Steinberger	8195497cec	fix: surface gateway slash commands in TUI	2026-01-23 18:58:41 +00:00
Peter Steinberger	1af227b619	fix: forward unknown TUI slash commands	2026-01-23 18:41:02 +00:00
Peter Steinberger	b77e730657	fix: add per-channel markdown table conversion (#1495 ) (thanks @odysseus0)	2026-01-23 18:39:25 +00:00
Peter Steinberger	37e5f077b8	test: move gateway server coverage to e2e	2026-01-23 18:34:33 +00:00
Peter Steinberger	0eb7e1864c	test: move auto-reply directive coverage to e2e	2026-01-23 18:34:33 +00:00
Peter Steinberger	0d336272f9	test: consolidate auto-reply unit coverage	2026-01-23 18:34:33 +00:00
Peter Steinberger	ace6a42ea6	test: dedupe CLI onboard auth cases	2026-01-23 18:34:33 +00:00
Peter Steinberger	6d2a1ce217	test: trim async waits in webhook tests	2026-01-23 18:34:33 +00:00
Peter Steinberger	c9d73469c3	test: stub heavy tools in agent tests	2026-01-23 18:34:33 +00:00
Peter Steinberger	29353e2e81	test: speed up default test env	2026-01-23 18:34:33 +00:00
Peter Steinberger	fdc50a0feb	fix: normalize session lock path	2026-01-23 18:34:33 +00:00
George Zhang	a1413a011e	feat(telegram): convert markdown tables to bullet points (#1495 ) Tables render poorly in Telegram (pipes stripped, whitespace collapses). This adds a 'tableMode' option to markdownToIR that converts tables to nested bullet points, which render cleanly on mobile. - Add tableMode: 'flat' \| 'bullets' to MarkdownParseOptions - Track table state during token rendering - Render tables as bullet points with first column as row labels - Apply bold styling to row labels for visual hierarchy - Enable tableMode: 'bullets' for Telegram formatter Closes #TBD	2026-01-23 18:00:51 +00:00
George Zhang	bfbeea0f20	daemon: prefer symlinked paths over realpath for stable service configs (#1505 ) When installing the LaunchAgent/systemd service, the CLI was using fs.realpath() to resolve the entry.js path, which converted stable symlinked paths (e.g. node_modules/clawdbot) into version-specific paths (e.g. .pnpm/clawdbot@X.Y.Z/...). This caused the service to break after pnpm updates because the old versioned path no longer exists, even though the symlink still works. Now we prefer the original (symlinked) path when it's valid, keeping service configs stable across package version updates.	2026-01-23 11:52:26 +00:00
Peter Steinberger	2c85b1b409	fix: restart gateway after update by default	2026-01-23 11:50:19 +00:00
Peter Steinberger	8b7b7e154f	chore: speed up tests and update opencode models	2026-01-23 11:36:32 +00:00
Peter Steinberger	bb9bddebb4	fix: stabilize ci tests	2026-01-23 09:52:22 +00:00
Peter Steinberger	6e570561b6	docs: prefer fast install smoke for release	2026-01-23 09:18:15 +00:00
Peter Steinberger	fb6363ae58	Merge pull request #1492 from svkozak/fix-discord-accountId Discord: preserve accountId in message actions (refs #1489)	2026-01-23 09:15:54 +00:00
Peter Steinberger	1b77e086d4	Merge origin/main into fix-discord-accountId	2026-01-23 09:15:44 +00:00
Sergii Kozak	d371a4c8c3	Discord Actions: Update tests for optional config parameter	2026-01-23 01:11:54 -08:00
Peter Steinberger	03e8b7c4ba	fix: always offer TUI hatch	2026-01-23 09:07:43 +00:00
Peter Steinberger	8aadcaa1bd	test: fix discord action mocks	2026-01-23 09:06:04 +00:00
Peter Steinberger	96800c27ec	docs: update changelog for #1492	2026-01-23 09:06:04 +00:00
Peter Steinberger	13d1712850	fix: honor accountId in message actions	2026-01-23 09:06:04 +00:00
Sergii Kozak	c5546f0d5b	Discord: preserve accountId in message actions (refs #1489 )	2026-01-23 09:06:04 +00:00
Peter Steinberger	3de5ea818d	ci: speed up install smoke on PRs	2026-01-23 09:05:15 +00:00
Peter Steinberger	dc07f1e021	fix: keep core tools when allowlist is plugin-only	2026-01-23 09:02:17 +00:00
Peter Steinberger	310a248a44	docs: add exe.dev ops note	2026-01-23 09:01:02 +00:00
Peter Steinberger	88e7684258	chore: update appcast for 2026.1.22	2026-01-23 08:59:04 +00:00
Sergii Kozak	716f901504	Discord: honor accountId across channel actions (refs #1489 )	2026-01-23 00:50:50 -08:00
Peter Steinberger	e817c0cee5	fix: preserve PNG alpha fallback (#1491 ) (thanks @robbyczgw-cla)	2026-01-23 08:45:50 +00:00
Robby	e634791585	fix(media): preserve alpha channel for transparent PNGs (#1473 )	2026-01-23 08:43:01 +00:00
Peter Steinberger	78071f8ec4	docs: note SPARKLE_PRIVATE_KEY_FILE in profile	2026-01-23 08:25:20 +00:00
Peter Steinberger	c48751a99c	chore: sync plugin versions for 2026.1.22	2026-01-23 08:18:55 +00:00
Peter Steinberger	86e0916fa3	fix: allow windows spawn in test parallel	2026-01-23 07:52:04 +00:00
Sergii Kozak	dc89bc4004	Discord: preserve accountId in message actions (refs #1489 )	2026-01-22 23:51:58 -08:00
Peter Steinberger	0c7e649676	docs: fix 2026.1.21 changelog placement	2026-01-23 07:51:40 +00:00
Peter Steinberger	45ce07a098	test: split vitest into unit and gateway	2026-01-23 07:34:57 +00:00
Peter Steinberger	aed8dc1ade	test: consolidate pi-tools shards	2026-01-23 07:34:57 +00:00
Peter Steinberger	86a341be62	test: speed up history and cron suites	2026-01-23 07:34:57 +00:00
Ian Hildebrand	ff78e9a564	fix: support direct token and provider in auth apply commands (#1485 )	2026-01-23 07:27:52 +00:00
Peter Steinberger	60a60779d7	test: streamline slow suites	2026-01-23 07:26:19 +00:00
Peter Steinberger	32da00cb2f	docs: note vitest worker cap	2026-01-23 07:26:19 +00:00
Peter Steinberger	0420f2804c	fix: log config update in copilot auth	2026-01-23 07:23:52 +00:00
Hiren PatelandPeter Steinberger	4de660bec6	[AI Assisted] Usage: add Google Antigravity usage tracking (#1490 ) * Usage: add Google Antigravity usage tracking - Add dedicated fetcher for google-antigravity provider - Fetch credits and per-model quotas from Cloud Code API - Report individual model IDs sorted by usage (top 10) - Include comprehensive debug logging with [antigravity] prefix * fix: refine antigravity usage tracking (#1490) (thanks @patelhiren) --------- Co-authored-by: Peter Steinberger <steipete@gmail.com>	2026-01-23 07:17:59 +00:00
Peter Steinberger	58f638463f	fix: stop gateway before uninstall	2026-01-23 07:17:42 +00:00
Peter Steinberger	f1afc722da	Revert "fix: improve GitHub Copilot integration" This reverts commit `21a9b3b66f`.	2026-01-23 07:14:00 +00:00
Peter Steinberger	bc75d58e9e	Revert "fix: set Copilot user agent header" This reverts commit `cfcc4548bb`.	2026-01-23 07:14:00 +00:00
Peter Steinberger	2efd265697	Revert "fix: treat copilot oauth tokens as non-expiring" This reverts commit `35228ecae9`.	2026-01-23 07:14:00 +00:00
Peter Steinberger	9c1f1476bc	docs: fix Lobster changelog placement	2026-01-23 07:12:13 +00:00
Peter Steinberger	e8352c8d21	fix: stabilize cron log wait	2026-01-23 07:11:01 +00:00
Peter Steinberger	551685351f	fix: sanitize assistant session text (#1456 ) (thanks @zerone0x)	2026-01-23 07:05:31 +00:00
Peter Steinberger	3fbbac07fe	fix: prioritize Anthropic token auth option	2026-01-23 07:04:18 +00:00
zerone0xandClaude	03bec49299	fix: sanitize tool call text in sessions-helpers extractAssistantText Adds sanitization to extractAssistantText in sessions-helpers.ts to prevent tool call text from leaking to users. Previously, messages retrieved from chat history via sessions-helpers.ts could expose: - Minimax XML tool calls (<invoke>...</invoke>) - Downgraded tool call markers ([Tool Call: name (ID: ...)]) - Thinking tags (<think>...</think>) This fix: - Exports the stripping functions from pi-embedded-utils.ts - Adds a new sanitizeTextContent helper in sessions-helpers.ts - Updates extractAssistantText to sanitize before returning - Updates extractMessageText in commands-subagents.ts to sanitize Fixes #1269 Co-Authored-By: Claude <noreply@anthropic.com>	2026-01-23 07:03:26 +00:00
Peter Steinberger	6779ba2367	fix(tui): hide off think/verbose in footer	2026-01-23 07:02:56 +00:00
Peter Steinberger	8598e906ef	docs: highlight compaction safeguards in changelog	2026-01-23 06:41:23 +00:00
Peter Steinberger	300fc486a4	test: avoid double cron finish wait	2026-01-23 06:40:14 +00:00
Peter Steinberger	f014b46b56	test: harden onboarding/discord/telegram test setup	2026-01-23 06:38:16 +00:00
Peter Steinberger	833f5acda1	test: stabilize cron + async search timings	2026-01-23 06:38:16 +00:00
Dave Lauer	d03c404cb4	feat(compaction): add adaptive chunk sizing, progressive fallback, and UI indicator (#1466 ) * fix(ui): allow relative URLs in avatar validation The isAvatarUrl check only accepted http://, https://, or data: URLs, but the /avatar/{agentId} endpoint returns relative paths like /avatar/main. This caused local file avatars to display as text instead of images. Fixes avatar display for locally configured avatar files. * fix(gateway): resolve local avatars to URL in HTML injection and RPC The frontend fix alone wasn't enough because: 1. serveIndexHtml() was injecting the raw avatar filename into HTML 2. agent.identity.get RPC was returning raw filename, overwriting the HTML-injected value Now both paths resolve local file avatars (.png, .jpg, etc.) to the /avatar/{agentId} endpoint URL. * feat(compaction): add adaptive chunk sizing and progressive fallback - Add computeAdaptiveChunkRatio() to reduce chunk size for large messages - Add isOversizedForSummary() to detect messages too large to summarize - Add summarizeWithFallback() with progressive fallback: - Tries full summarization first - Falls back to partial summarization excluding oversized messages - Notes oversized messages in the summary output - Add SAFETY_MARGIN (1.2x) buffer for token estimation inaccuracy - Reduce MIN_CHUNK_RATIO to 0.15 for very large messages This prevents compaction failures when conversations contain unusually large tool outputs or responses that exceed the summarization model's context window. * feat(ui): add compaction indicator and improve event error handling Compaction indicator: - Add CompactionStatus type and handleCompactionEvent() in app-tool-stream.ts - Show '🧹 Compacting context...' toast while active (with pulse animation) - Show '🧹 Context compacted' briefly after completion - Auto-clear toast after 5 seconds - Add CSS styles for .callout.info, .callout.success, .compaction-indicator Error handling improvements: - Wrap onEvent callback in try/catch in gateway.ts to prevent errors from breaking the WebSocket message handler - Wrap handleGatewayEvent in try/catch with console.error logging to isolate errors and make them visible in devtools These changes address UI freezes during heavy agent activity by: 1. Showing users when compaction is happening 2. Preventing uncaught errors from silently breaking the event loop * fix(control-ui): add agentId to DEFAULT_ASSISTANT_IDENTITY TypeScript inferred the union type without agentId when falling back to DEFAULT_ASSISTANT_IDENTITY, causing build errors at control-ui.ts:222-223.	2026-01-23 06:32:30 +00:00
Peter Steinberger	68ea6e521b	fix: reduce Slack WebClient retries	2026-01-23 06:31:53 +00:00
Peter Steinberger	4912e85ac8	fix: fall back to non-PTY exec	2026-01-23 06:27:26 +00:00
Peter Steinberger	39d8ff59aa	test: trim plugin + telegram test setup	2026-01-23 06:22:09 +00:00
Peter Steinberger	070944f64f	test(memory): speed up batch coverage	2026-01-23 06:22:09 +00:00
Peter Steinberger	d4db45e8a9	test(agents): merge sessions_spawn group announce coverage	2026-01-23 06:22:09 +00:00
Peter Steinberger	451792d326	test(commands): streamline onboarding tests	2026-01-23 06:22:09 +00:00
Peter Steinberger	c7ca312f97	test(gateway): consolidate server suites for speed	2026-01-23 06:22:09 +00:00
ganghyun kimandGanghyun Kim	1e6e58b23b	fix: clarify Discord onboarding hint (#1487 ) Thanks @kyleok. Co-authored-by: Ganghyun Kim <58307870+kyleok@users.noreply.github.com>	2026-01-23 06:11:41 +00:00
Peter Steinberger	e98e71401a	fix: always skip browser opens in tests	2026-01-23 06:00:21 +00:00
Peter Steinberger	bec1d0d3d4	fix: extend gateway chat test timeout on windows	2026-01-23 05:55:35 +00:00
Peter Steinberger	9f6ea67415	fix: gateway summary lookup + test browser opens	2026-01-23 05:54:51 +00:00
Peter Steinberger	bd7443b39b	docs: update media auto-detect	2026-01-23 05:47:16 +00:00
Peter Steinberger	93bef830ce	test: add media auto-detect coverage	2026-01-23 05:47:13 +00:00
Peter Steinberger	2dfbd1c1f6	feat: improve media auto-detect	2026-01-23 05:47:09 +00:00
Peter Steinberger	1d9f230be4	docs: expand slack replyToModeByChatType examples	2026-01-23 05:38:28 +00:00
Peter Steinberger	9bf295da48	feat: add slack replyToModeByChatType overrides	2026-01-23 05:38:28 +00:00
Peter Steinberger	eebd750781	fix: improve matrix direct room resolution (#1436 ) (thanks @sibbl) (#1486 ) * fix: improve matrix direct room resolution (#1436) (thanks @sibbl) * docs: update changelog for matrix fix (#1486) (thanks @sibbl)	2026-01-23 05:38:04 +00:00
Sebastian Schubotz	aa11300175	fix(matrix): broken import and enhance direct room resolve logic (#1436 ) * fix(matrix): fix broken import again * fix(matrix): improve error handling and fallback logic in resolveDirectRoomId	2026-01-23 05:35:01 +00:00