fix: improve sandbox image inspect handling (#1548) (thanks @sweepies)

Updated `dockerImageExists` in `src/commands/doctor-sandbox.ts` to mirror the logic in `src/agents/sandbox/docker.ts`. It now re-throws errors unless they are explicitly "No such image" errors.

AGENTS.md

@@ -23,13 +23,14 @@
 - 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:
+- Access: stable path is `ssh exe.dev` then `ssh vm-name` (assume SSH key already set).
+- SSH flaky: use exe.dev web terminal or Shelley (web agent); keep a tmux session for long ops.
+- Update: `sudo npm i -g clawdbot@latest` (global install needs root on `/usr/lib/node_modules`).
+- Config: use `clawdbot config set ...`; ensure `gateway.mode=local` is set.
+- Discord: store raw token only (no `DISCORD_BOT_TOKEN=` prefix).
+- Restart: 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.
+- Verify: `clawdbot channels status --probe`, `ss -ltnp | rg 18789`, `tail -n 120 /tmp/clawdbot-gateway.log`.
 
 ## Build, Test, and Development Commands
 - Runtime baseline: Node **22+** (keep Node + Bun paths working).

CHANGELOG.md

@@ -2,43 +2,24 @@
 
 Docs: https://docs.clawd.bot
 
-## 2026.1.23 (Unreleased)
+## 2026.1.23
 
 ### Changes
-- Plugins: add optional llm-task JSON-only tool for workflows. (#1498) Thanks @vignesh07.
 - 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.
-- Agents: add Bedrock auto-discovery defaults + config overrides. (#1553) Thanks @fal3.
-- Docs: add cron vs heartbeat decision guide (with Lobster workflow notes). (#1533) Thanks @JustYannicc.
-- Docs: clarify HEARTBEAT.md empty file skips heartbeats, missing file still runs. (#1535) Thanks @JustYannicc.
 - Markdown: add per-channel table conversion (bullets for Signal/WhatsApp, code blocks elsewhere). (#1495) Thanks @odysseus0.
-- Tlon: add Urbit channel plugin (DMs, group mentions, thread replies). (#1544) Thanks @wca4a.
 
 ### Fixes
-- Docker: update gateway command in docker-compose and Hetzner guide. (#1514)
 - Voice wake: auto-save wake words on blur/submit across iOS/Android and align limits with macOS.
-- UI: keep the Control UI sidebar visible while scrolling long pages. (#1515) Thanks @pookNast.
-- UI: cache Control UI markdown rendering + memoize chat text extraction to reduce Safari typing jank.
-- Tailscale: retry serve/funnel with sudo only for permission errors and keep original failure details. (#1551) Thanks @sweepies.
-- Agents: add CLI log hint to "agent failed before reply" messages. (#1550) Thanks @sweepies.
 - Discord: limit autoThread mention bypass to bot-owned threads; keep ack reactions mention-gated. (#1511) Thanks @pvoo.
 - Gateway: accept null optional fields in exec approval requests. (#1511) Thanks @pvoo.
+- Sandbox: log docker image inspect failures during doctor checks. (#1548) Thanks @sweepies.
 - 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.
 - CLI: hide auth probe timeout warnings from embedded runs.
 - CLI: render auth probe results as a table in `clawdbot models status`.
-- CLI: suppress probe-only embedded logs unless `--verbose` is set.
-- CLI: move auth probe errors below the table to reduce wrapping.
-- CLI: prevent ANSI color bleed when table cells wrap.
-- CLI: explain when auth profiles are excluded by auth.order in probe details.
-- CLI: drop the em dash when the banner tagline wraps to a second line.
-- CLI: inline auth probe errors in status rows to reduce wrapping.
-- Telegram: render markdown in media captions. (#1478)
-- Agents: honor enqueue overrides for embedded runs to avoid queue deadlocks in tests.
-- Daemon: use platform PATH delimiters when building minimal service paths.
-- Tests: skip embedded runner ordering assertion on Windows to avoid CI timeouts.
 - 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.

apps/android/app/build.gradle.kts

@@ -21,8 +21,8 @@ android {
     applicationId = "com.clawdbot.android"
     minSdk = 31
     targetSdk = 36
-    versionCode = 202601210
-    versionName = "2026.1.21"
+    versionCode = 202601230
+    versionName = "2026.1.23"
   }
 
   buildTypes {

apps/android/app/src/main/java/com/clawdbot/android/WakeWords.kt

@@ -8,10 +8,14 @@ object WakeWords {
     return input.split(",").map { it.trim() }.filter { it.isNotEmpty() }
   }
 
+  fun parseIfChanged(input: String, current: List<String>): List<String>? {
+    val parsed = parseCommaSeparated(input)
+    return if (parsed == current) null else parsed
+  }
+
   fun sanitize(words: List<String>, defaults: List<String>): List<String> {
     val cleaned =
       words.map { it.trim() }.filter { it.isNotEmpty() }.take(maxWords).map { it.take(maxWordLength) }
     return cleaned.ifEmpty { defaults }
   }
 }
-

apps/android/app/src/main/java/com/clawdbot/android/ui/SettingsSheet.kt

@@ -28,6 +28,8 @@ import androidx.compose.foundation.layout.windowInsetsPadding
 import androidx.compose.foundation.lazy.LazyColumn
 import androidx.compose.foundation.lazy.items
 import androidx.compose.foundation.lazy.rememberLazyListState
+import androidx.compose.foundation.text.KeyboardActions
+import androidx.compose.foundation.text.KeyboardOptions
 import androidx.compose.material.icons.Icons
 import androidx.compose.material.icons.filled.ExpandLess
 import androidx.compose.material.icons.filled.ExpandMore
@@ -49,7 +51,10 @@ import androidx.compose.runtime.remember
 import androidx.compose.runtime.setValue
 import androidx.compose.ui.Modifier
 import androidx.compose.ui.draw.alpha
+import androidx.compose.ui.focus.onFocusChanged
 import androidx.compose.ui.platform.LocalContext
+import androidx.compose.ui.platform.LocalFocusManager
+import androidx.compose.ui.text.input.ImeAction
 import androidx.compose.ui.text.style.TextAlign
 import androidx.compose.ui.unit.dp
 import androidx.core.content.ContextCompat
@@ -58,6 +63,7 @@ import com.clawdbot.android.LocationMode
 import com.clawdbot.android.MainViewModel
 import com.clawdbot.android.NodeForegroundService
 import com.clawdbot.android.VoiceWakeMode
+import com.clawdbot.android.WakeWords
 
 @Composable
 fun SettingsSheet(viewModel: MainViewModel) {
@@ -86,6 +92,8 @@ fun SettingsSheet(viewModel: MainViewModel) {
   val listState = rememberLazyListState()
   val (wakeWordsText, setWakeWordsText) = remember { mutableStateOf("") }
   val (advancedExpanded, setAdvancedExpanded) = remember { mutableStateOf(false) }
+  val focusManager = LocalFocusManager.current
+  var wakeWordsHadFocus by remember { mutableStateOf(false) }
   val deviceModel =
     remember {
       listOfNotNull(Build.MANUFACTURER, Build.MODEL)
@@ -104,6 +112,12 @@ fun SettingsSheet(viewModel: MainViewModel) {
     }
 
   LaunchedEffect(wakeWords) { setWakeWordsText(wakeWords.joinToString(", ")) }
+  val commitWakeWords = {
+    val parsed = WakeWords.parseIfChanged(wakeWordsText, wakeWords)
+    if (parsed != null) {
+      viewModel.setWakeWords(parsed)
+    }
+  }
 
   val permissionLauncher =
     rememberLauncherForActivityResult(ActivityResultContracts.RequestMultiplePermissions()) { perms ->
@@ -481,25 +495,27 @@ fun SettingsSheet(viewModel: MainViewModel) {
         value = wakeWordsText,
         onValueChange = setWakeWordsText,
         label = { Text("Wake Words (comma-separated)") },
-        modifier = Modifier.fillMaxWidth(),
+        modifier =
+          Modifier.fillMaxWidth().onFocusChanged { focusState ->
+            if (focusState.isFocused) {
+              wakeWordsHadFocus = true
+            } else if (wakeWordsHadFocus) {
+              wakeWordsHadFocus = false
+              commitWakeWords()
+            }
+          },
         singleLine = true,
+        keyboardOptions = KeyboardOptions(imeAction = ImeAction.Done),
+        keyboardActions =
+          KeyboardActions(
+            onDone = {
+              commitWakeWords()
+              focusManager.clearFocus()
+            },
+          ),
       )
     }
-    item {
-      Row(horizontalArrangement = Arrangement.spacedBy(12.dp)) {
-        Button(
-          onClick = {
-            val parsed = com.clawdbot.android.WakeWords.parseCommaSeparated(wakeWordsText)
-            viewModel.setWakeWords(parsed)
-          },
-          enabled = isConnected,
-        ) {
-          Text("Save + Sync")
-        }
-
-        Button(onClick = viewModel::resetWakeWordsDefaults) { Text("Reset defaults") }
-      }
-    }
+    item { Button(onClick = viewModel::resetWakeWordsDefaults) { Text("Reset defaults") } }
     item {
       Text(
         if (isConnected) {

apps/android/app/src/test/java/com/clawdbot/android/WakeWordsTest.kt

@@ -1,6 +1,7 @@
 package com.clawdbot.android
 
 import org.junit.Assert.assertEquals
+import org.junit.Assert.assertNull
 import org.junit.Test
 
 class WakeWordsTest {
@@ -32,5 +33,18 @@ class WakeWordsTest {
     assertEquals("w1", sanitized.first())
     assertEquals("w${WakeWords.maxWords}", sanitized.last())
   }
-}
 
+  @Test
+  fun parseIfChangedSkipsWhenUnchanged() {
+    val current = listOf("clawd", "claude")
+    val parsed = WakeWords.parseIfChanged(" clawd , claude ", current)
+    assertNull(parsed)
+  }
+
+  @Test
+  fun parseIfChangedReturnsUpdatedList() {
+    val current = listOf("clawd")
+    val parsed = WakeWords.parseIfChanged(" clawd , jarvis ", current)
+    assertEquals(listOf("clawd", "jarvis"), parsed)
+  }
+}

apps/ios/Sources/Info.plist

@@ -19,9 +19,9 @@
 	<key>CFBundlePackageType</key>
 	<string>APPL</string>
 	<key>CFBundleShortVersionString</key>
-	<string>2026.1.21</string>
+	<string>2026.1.23</string>
 	<key>CFBundleVersion</key>
-	<string>20260121</string>
+	<string>20260123</string>
 	<key>NSAppTransportSecurity</key>
 	<dict>
 		<key>NSAllowsArbitraryLoadsInWebContent</key>

apps/ios/Sources/Settings/VoiceWakeWordsSettingsView.swift

@@ -1,8 +1,10 @@
 import SwiftUI
+import Combine
 
 struct VoiceWakeWordsSettingsView: View {
     @Environment(NodeAppModel.self) private var appModel
     @State private var triggerWords: [String] = VoiceWakePreferences.loadTriggerWords()
+    @FocusState private var focusedTriggerIndex: Int?
     @State private var syncTask: Task<Void, Never>?
 
     var body: some View {
@@ -12,6 +14,10 @@ struct VoiceWakeWordsSettingsView: View {
                     TextField("Wake word", text: self.binding(for: index))
                         .textInputAutocapitalization(.never)
                         .autocorrectionDisabled()
+                        .focused(self.$focusedTriggerIndex, equals: index)
+                        .onSubmit {
+                            self.commitTriggerWords()
+                        }
                 }
                 .onDelete(perform: self.removeWords)
 
@@ -39,17 +45,18 @@ struct VoiceWakeWordsSettingsView: View {
         .onAppear {
             if self.triggerWords.isEmpty {
                 self.triggerWords = VoiceWakePreferences.defaultTriggerWords
+                self.commitTriggerWords()
             }
         }
-        .onChange(of: self.triggerWords) { _, newValue in
-            // Keep local voice wake responsive even if the gateway isn't connected yet.
-            VoiceWakePreferences.saveTriggerWords(newValue)
-
-            let snapshot = VoiceWakePreferences.sanitizeTriggerWords(newValue)
-            self.syncTask?.cancel()
-            self.syncTask = Task { [snapshot, weak appModel = self.appModel] in
-                try? await Task.sleep(nanoseconds: 650_000_000)
-                await appModel?.setGlobalWakeWords(snapshot)
+        .onChange(of: self.focusedTriggerIndex) { oldValue, newValue in
+            guard oldValue != nil, oldValue != newValue else { return }
+            self.commitTriggerWords()
+        }
+        .onReceive(NotificationCenter.default.publisher(for: UserDefaults.didChangeNotification)) { _ in
+            guard self.focusedTriggerIndex == nil else { return }
+            let updated = VoiceWakePreferences.loadTriggerWords()
+            if updated != self.triggerWords {
+                self.triggerWords = updated
             }
         }
     }
@@ -63,6 +70,7 @@ struct VoiceWakeWordsSettingsView: View {
         if self.triggerWords.isEmpty {
             self.triggerWords = VoiceWakePreferences.defaultTriggerWords
         }
+        self.commitTriggerWords()
     }
 
     private func binding(for index: Int) -> Binding<String> {
@@ -76,4 +84,15 @@ struct VoiceWakeWordsSettingsView: View {
                 self.triggerWords[index] = newValue
             })
     }
+
+    private func commitTriggerWords() {
+        VoiceWakePreferences.saveTriggerWords(self.triggerWords)
+
+        let snapshot = VoiceWakePreferences.sanitizeTriggerWords(self.triggerWords)
+        self.syncTask?.cancel()
+        self.syncTask = Task { [snapshot, weak appModel = self.appModel] in
+            try? await Task.sleep(nanoseconds: 650_000_000)
+            await appModel?.setGlobalWakeWords(snapshot)
+        }
+    }
 }

apps/ios/Sources/Voice/VoiceWakePreferences.swift

@@ -6,6 +6,8 @@ enum VoiceWakePreferences {
 
     // Keep defaults aligned with the mac app.
     static let defaultTriggerWords: [String] = ["clawd", "claude"]
+    static let maxWords = 32
+    static let maxWordLength = 64
 
     static func decodeGatewayTriggers(from payloadJSON: String) -> [String]? {
         guard let data = payloadJSON.data(using: .utf8) else { return nil }
@@ -30,6 +32,8 @@ enum VoiceWakePreferences {
         let cleaned = words
             .map { $0.trimmingCharacters(in: .whitespacesAndNewlines) }
             .filter { !$0.isEmpty }
+            .prefix(Self.maxWords)
+            .map { String($0.prefix(Self.maxWordLength)) }
         return cleaned.isEmpty ? Self.defaultTriggerWords : cleaned
     }
 

apps/ios/Tests/Info.plist

@@ -17,8 +17,8 @@
 	<key>CFBundlePackageType</key>
 	<string>BNDL</string>
 	<key>CFBundleShortVersionString</key>
-	<string>2026.1.21</string>
+	<string>2026.1.23</string>
 	<key>CFBundleVersion</key>
-	<string>20260121</string>
+	<string>20260123</string>
 </dict>
 </plist>

apps/ios/Tests/VoiceWakePreferencesTests.swift

@@ -11,6 +11,18 @@ import Testing
         #expect(VoiceWakePreferences.sanitizeTriggerWords(["", "  "]) == VoiceWakePreferences.defaultTriggerWords)
     }
 
+    @Test func sanitizeTriggerWordsLimitsWordLength() {
+        let long = String(repeating: "x", count: VoiceWakePreferences.maxWordLength + 5)
+        let cleaned = VoiceWakePreferences.sanitizeTriggerWords(["ok", long])
+        #expect(cleaned[1].count == VoiceWakePreferences.maxWordLength)
+    }
+
+    @Test func sanitizeTriggerWordsLimitsWordCount() {
+        let words = (1...VoiceWakePreferences.maxWords + 3).map { "w\($0)" }
+        let cleaned = VoiceWakePreferences.sanitizeTriggerWords(words)
+        #expect(cleaned.count == VoiceWakePreferences.maxWords)
+    }
+
     @Test func displayStringUsesSanitizedWords() {
         #expect(VoiceWakePreferences.displayString(for: ["", " "]) == "clawd, claude")
     }

apps/ios/project.yml

@@ -81,8 +81,8 @@ targets:
       properties:
         CFBundleDisplayName: Clawdbot
         CFBundleIconName: AppIcon
-        CFBundleShortVersionString: "2026.1.21"
-        CFBundleVersion: "20260121"
+        CFBundleShortVersionString: "2026.1.23"
+        CFBundleVersion: "20260123"
         UILaunchScreen: {}
         UIApplicationSceneManifest:
           UIApplicationSupportsMultipleScenes: false
@@ -130,5 +130,5 @@ targets:
       path: Tests/Info.plist
       properties:
         CFBundleDisplayName: ClawdbotTests
-        CFBundleShortVersionString: "2026.1.21"
-        CFBundleVersion: "20260121"
+        CFBundleShortVersionString: "2026.1.23"
+        CFBundleVersion: "20260123"

apps/macos/Sources/Clawdbot/Constants.swift

@@ -12,6 +12,8 @@ let voiceWakeTriggerChimeKey = "clawdbot.voiceWakeTriggerChime"
 let voiceWakeSendChimeKey = "clawdbot.voiceWakeSendChime"
 let showDockIconKey = "clawdbot.showDockIcon"
 let defaultVoiceWakeTriggers = ["clawd", "claude"]
+let voiceWakeMaxWords = 32
+let voiceWakeMaxWordLength = 64
 let voiceWakeMicKey = "clawdbot.voiceWakeMicID"
 let voiceWakeMicNameKey = "clawdbot.voiceWakeMicName"
 let voiceWakeLocaleKey = "clawdbot.voiceWakeLocaleID"

apps/macos/Sources/Clawdbot/Resources/Info.plist

@@ -15,9 +15,9 @@
     <key>CFBundlePackageType</key>
     <string>APPL</string>
     <key>CFBundleShortVersionString</key>
-    <string>2026.1.21</string>
+    <string>2026.1.23</string>
     <key>CFBundleVersion</key>
-    <string>202601210</string>
+    <string>202601230</string>
     <key>CFBundleIconFile</key>
     <string>Clawdbot</string>
     <key>CFBundleURLTypes</key>

apps/macos/Sources/Clawdbot/VoiceWakeHelpers.swift

@@ -4,6 +4,8 @@ func sanitizeVoiceWakeTriggers(_ words: [String]) -> [String] {
     let cleaned = words
         .map { $0.trimmingCharacters(in: .whitespacesAndNewlines) }
         .filter { !$0.isEmpty }
+        .prefix(voiceWakeMaxWords)
+        .map { String($0.prefix(voiceWakeMaxWordLength)) }
     return cleaned.isEmpty ? defaultVoiceWakeTriggers : cleaned
 }
 

apps/macos/Tests/ClawdbotIPCTests/VoiceWakeHelpersTests.swift

@@ -12,6 +12,18 @@ struct VoiceWakeHelpersTests {
         #expect(cleaned == defaultVoiceWakeTriggers)
     }
 
+    @Test func sanitizeTriggersLimitsWordLength() {
+        let long = String(repeating: "x", count: voiceWakeMaxWordLength + 5)
+        let cleaned = sanitizeVoiceWakeTriggers(["ok", long])
+        #expect(cleaned[1].count == voiceWakeMaxWordLength)
+    }
+
+    @Test func sanitizeTriggersLimitsWordCount() {
+        let words = (1...voiceWakeMaxWords + 3).map { "w\($0)" }
+        let cleaned = sanitizeVoiceWakeTriggers(words)
+        #expect(cleaned.count == voiceWakeMaxWords)
+    }
+
     @Test func normalizeLocaleStripsCollation() {
         #expect(normalizeLocaleIdentifier("en_US@collation=phonebook") == "en_US")
     }

docs/gateway/heartbeat.md

@@ -159,10 +159,6 @@ If a `HEARTBEAT.md` file exists in the workspace, the default prompt tells the
 agent to read it. Think of it as your “heartbeat checklist”: small, stable, and
 safe to include every 30 minutes.
 
-If `HEARTBEAT.md` exists but is effectively empty (only blank lines and markdown
-headers like `# Heading`), Clawdbot skips the heartbeat run to save API calls.
-If the file is missing, the heartbeat still runs and the model decides what to do.
-
 Keep it tiny (short checklist or reminders) to avoid prompt bloat.
 
 Example `HEARTBEAT.md`:

docs/help/troubleshooting.md

@@ -43,6 +43,14 @@ Almost always a Node/npm PATH issue. Start here:
 - [Gateway troubleshooting](/gateway/troubleshooting)
 - [Control UI](/web/control-ui#insecure-http)
 
+### `docs.clawd.bot` shows an SSL error (Comcast/Xfinity)
+
+Some Comcast/Xfinity connections block `docs.clawd.bot` via Xfinity Advanced Security.
+Disable Advanced Security or add `docs.clawd.bot` to the allowlist, then retry.
+
+- Xfinity Advanced Security help: https://www.xfinity.com/support/articles/using-xfinity-xfi-advanced-security
+- Quick sanity checks: try a mobile hotspot or VPN to confirm it’s ISP-level filtering
+
 ### Service says running, but RPC probe fails
 
 - [Gateway troubleshooting](/gateway/troubleshooting)

docs/platforms/mac/release.md

@@ -30,17 +30,17 @@ Notes:
 # From repo root; set release IDs so Sparkle feed is enabled.
 # APP_BUILD must be numeric + monotonic for Sparkle compare.
 BUNDLE_ID=com.clawdbot.mac \
-APP_VERSION=2026.1.21 \
+APP_VERSION=2026.1.23 \
 APP_BUILD="$(git rev-list --count HEAD)" \
 BUILD_CONFIG=release \
 SIGN_IDENTITY="Developer ID Application: <Developer Name> (<TEAMID>)" \
 scripts/package-mac-app.sh
 
 # Zip for distribution (includes resource forks for Sparkle delta support)
-ditto -c -k --sequesterRsrc --keepParent dist/Clawdbot.app dist/Clawdbot-2026.1.21.zip
+ditto -c -k --sequesterRsrc --keepParent dist/Clawdbot.app dist/Clawdbot-2026.1.23.zip
 
 # Optional: also build a styled DMG for humans (drag to /Applications)
-scripts/create-dmg.sh dist/Clawdbot.app dist/Clawdbot-2026.1.21.dmg
+scripts/create-dmg.sh dist/Clawdbot.app dist/Clawdbot-2026.1.23.dmg
 
 # Recommended: build + notarize/staple zip + DMG
 # First, create a keychain profile once:
@@ -48,26 +48,26 @@ scripts/create-dmg.sh dist/Clawdbot.app dist/Clawdbot-2026.1.21.dmg
 #     --apple-id "<apple-id>" --team-id "<team-id>" --password "<app-specific-password>"
 NOTARIZE=1 NOTARYTOOL_PROFILE=clawdbot-notary \
 BUNDLE_ID=com.clawdbot.mac \
-APP_VERSION=2026.1.21 \
+APP_VERSION=2026.1.23 \
 APP_BUILD="$(git rev-list --count HEAD)" \
 BUILD_CONFIG=release \
 SIGN_IDENTITY="Developer ID Application: <Developer Name> (<TEAMID>)" \
 scripts/package-mac-dist.sh
 
 # Optional: ship dSYM alongside the release
-ditto -c -k --keepParent apps/macos/.build/release/Clawdbot.app.dSYM dist/Clawdbot-2026.1.21.dSYM.zip
+ditto -c -k --keepParent apps/macos/.build/release/Clawdbot.app.dSYM dist/Clawdbot-2026.1.23.dSYM.zip
 ```
 
 ## Appcast entry
 Use the release note generator so Sparkle renders formatted HTML notes:
 ```bash
-SPARKLE_PRIVATE_KEY_FILE=/path/to/ed25519-private-key scripts/make_appcast.sh dist/Clawdbot-2026.1.21.zip https://raw.githubusercontent.com/clawdbot/clawdbot/main/appcast.xml
+SPARKLE_PRIVATE_KEY_FILE=/path/to/ed25519-private-key scripts/make_appcast.sh dist/Clawdbot-2026.1.23.zip https://raw.githubusercontent.com/clawdbot/clawdbot/main/appcast.xml
 ```
 Generates HTML release notes from `CHANGELOG.md` (via [`scripts/changelog-to-html.sh`](https://github.com/clawdbot/clawdbot/blob/main/scripts/changelog-to-html.sh)) and embeds them in the appcast entry.
 Commit the updated `appcast.xml` alongside the release assets (zip + dSYM) when publishing.
 
 ## Publish & verify
-- Upload `Clawdbot-2026.1.21.zip` (and `Clawdbot-2026.1.21.dSYM.zip`) to the GitHub release for tag `v2026.1.21`.
+- Upload `Clawdbot-2026.1.23.zip` (and `Clawdbot-2026.1.23.dSYM.zip`) to the GitHub release for tag `v2026.1.23`.
 - Ensure the raw appcast URL matches the baked feed: `https://raw.githubusercontent.com/clawdbot/clawdbot/main/appcast.xml`.
 - Sanity checks:
   - `curl -I https://raw.githubusercontent.com/clawdbot/clawdbot/main/appcast.xml` returns 200.

docs/reference/templates/HEARTBEAT.md

@@ -5,5 +5,4 @@ read_when:
 ---
 # HEARTBEAT.md
 
-# Keep this file empty (or with only comments) to skip heartbeat API calls.
-# Add tasks below when you want the agent to check something periodically.
+Keep this file empty unless you want a tiny checklist. Keep it small.

docs/start/clawd.md

@@ -182,8 +182,6 @@ By default, Clawdbot runs a heartbeat every 30 minutes with the prompt:
 `Read HEARTBEAT.md if it exists (workspace context). Follow it strictly. Do not infer or repeat old tasks from prior chats. If nothing needs attention, reply HEARTBEAT_OK.`
 Set `agents.defaults.heartbeat.every: "0m"` to disable.
 
-- If `HEARTBEAT.md` exists but is effectively empty (only blank lines and markdown headers like `# Heading`), Clawdbot skips the heartbeat run to save API calls.
-- If the file is missing, the heartbeat still runs and the model decides what to do.
 - If the agent replies with `HEARTBEAT_OK` (optionally with short padding; see `agents.defaults.heartbeat.ackMaxChars`), Clawdbot suppresses outbound delivery for that heartbeat.
 - Heartbeats run full agent turns — shorter intervals burn more tokens.
 

docs/start/faq.md

@@ -971,10 +971,6 @@ Heartbeats run every **30m** by default. Tune or disable them:
 }
 ```
 
-If `HEARTBEAT.md` exists but is effectively empty (only blank lines and markdown
-headers like `# Heading`), Clawdbot skips the heartbeat run to save API calls.
-If the file is missing, the heartbeat still runs and the model decides what to do.
-
 Per-agent overrides use `agents.list[].heartbeat`. Docs: [Heartbeat](/gateway/heartbeat).
 
 ### Do I need to add a “bot account” to a WhatsApp group?

extensions/bluebubbles/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clawdbot/bluebubbles",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "Clawdbot BlueBubbles channel plugin",
   "clawdbot": {

extensions/bluebubbles/src/monitor.test.ts

@@ -2,6 +2,7 @@ import { afterEach, beforeEach, describe, expect, it, vi } from "vitest";
 import type { IncomingMessage, ServerResponse } from "node:http";
 import { EventEmitter } from "node:events";
 
+import { removeAckReactionAfterReply, shouldAckReaction } from "clawdbot/plugin-sdk";
 import type { ClawdbotConfig, PluginRuntime } from "clawdbot/plugin-sdk";
 import {
   handleBlueBubblesWebhookRequest,
@@ -128,6 +129,7 @@ function createMockRuntime(): PluginRuntime {
       session: {
         resolveStorePath: mockResolveStorePath as unknown as PluginRuntime["channel"]["session"]["resolveStorePath"],
         readSessionUpdatedAt: mockReadSessionUpdatedAt as unknown as PluginRuntime["channel"]["session"]["readSessionUpdatedAt"],
+        recordInboundSession: vi.fn() as unknown as PluginRuntime["channel"]["session"]["recordInboundSession"],
         recordSessionMetaFromInbound: vi.fn() as unknown as PluginRuntime["channel"]["session"]["recordSessionMetaFromInbound"],
         updateLastRoute: vi.fn() as unknown as PluginRuntime["channel"]["session"]["updateLastRoute"],
       },
@@ -135,6 +137,10 @@ function createMockRuntime(): PluginRuntime {
         buildMentionRegexes: mockBuildMentionRegexes as unknown as PluginRuntime["channel"]["mentions"]["buildMentionRegexes"],
         matchesMentionPatterns: mockMatchesMentionPatterns as unknown as PluginRuntime["channel"]["mentions"]["matchesMentionPatterns"],
       },
+      reactions: {
+        shouldAckReaction,
+        removeAckReactionAfterReply,
+      },
       groups: {
         resolveGroupPolicy: mockResolveGroupPolicy as unknown as PluginRuntime["channel"]["groups"]["resolveGroupPolicy"],
         resolveRequireMention: mockResolveRequireMention as unknown as PluginRuntime["channel"]["groups"]["resolveRequireMention"],

extensions/bluebubbles/src/monitor.ts

@@ -1,7 +1,13 @@
 import type { IncomingMessage, ServerResponse } from "node:http";
 
 import type { ClawdbotConfig } from "clawdbot/plugin-sdk";
-import { resolveAckReaction } from "clawdbot/plugin-sdk";
+import {
+  logAckFailure,
+  logInboundDrop,
+  logTypingFailure,
+  resolveAckReaction,
+  resolveControlCommandGate,
+} from "clawdbot/plugin-sdk";
 import { markBlueBubblesChatRead, sendBlueBubblesTyping } from "./chat.js";
 import { resolveChatGuidForTarget, sendMessageBlueBubbles } from "./send.js";
 import { downloadBlueBubblesAttachment } from "./attachments.js";
@@ -1346,23 +1352,25 @@ async function processMessage(
         })
       : false;
   const dmAuthorized = dmPolicy === "open" || ownerAllowedForCommands;
-  const commandAuthorized = isGroup
-    ? core.channel.commands.resolveCommandAuthorizedFromAuthorizers({
-        useAccessGroups,
-        authorizers: [
-          { configured: effectiveAllowFrom.length > 0, allowed: ownerAllowedForCommands },
-          { configured: effectiveGroupAllowFrom.length > 0, allowed: groupAllowedForCommands },
-        ],
-      })
-    : dmAuthorized;
+  const commandGate = resolveControlCommandGate({
+    useAccessGroups,
+    authorizers: [
+      { configured: effectiveAllowFrom.length > 0, allowed: ownerAllowedForCommands },
+      { configured: effectiveGroupAllowFrom.length > 0, allowed: groupAllowedForCommands },
+    ],
+    allowTextCommands: true,
+    hasControlCommand: hasControlCmd,
+  });
+  const commandAuthorized = isGroup ? commandGate.commandAuthorized : dmAuthorized;
 
   // Block control commands from unauthorized senders in groups
-  if (isGroup && hasControlCmd && !commandAuthorized) {
-    logVerbose(
-      core,
-      runtime,
-      `bluebubbles: drop control command from unauthorized sender ${message.senderId}`,
-    );
+  if (isGroup && commandGate.shouldBlock) {
+    logInboundDrop({
+      log: (msg) => logVerbose(core, runtime, msg),
+      channel: "bluebubbles",
+      reason: "control command (unauthorized)",
+      target: message.senderId,
+    });
     return;
   }
 
@@ -1521,19 +1529,20 @@ async function processMessage(
     core,
     runtime,
   });
-  const shouldAckReaction = () => {
-    if (!ackReactionValue) return false;
-    if (ackReactionScope === "all") return true;
-    if (ackReactionScope === "direct") return !isGroup;
-    if (ackReactionScope === "group-all") return isGroup;
-    if (ackReactionScope === "group-mentions") {
-      if (!isGroup) return false;
-      if (!requireMention) return false;
-      if (!canDetectMention) return false;
-      return effectiveWasMentioned;
-    }
-    return false;
-  };
+  const shouldAckReaction = () =>
+    Boolean(
+      ackReactionValue &&
+        core.channel.reactions.shouldAckReaction({
+          scope: ackReactionScope,
+          isDirect: !isGroup,
+          isGroup,
+          isMentionableGroup: isGroup,
+          requireMention: Boolean(requireMention),
+          canDetectMention,
+          effectiveWasMentioned,
+          shouldBypassMention,
+        }),
+    );
   const ackMessageId = message.messageId?.trim() || "";
   const ackReactionPromise =
     shouldAckReaction() && ackMessageId && chatGuidForActions && ackReactionValue
@@ -1749,29 +1758,27 @@ async function processMessage(
       },
     });
   } finally {
-    if (
-      removeAckAfterReply &&
-      sentMessage &&
-      ackReactionPromise &&
-      ackReactionValue &&
-      chatGuidForActions &&
-      ackMessageId
-    ) {
-      void ackReactionPromise.then((didAck) => {
-        if (!didAck) return;
-        sendBlueBubblesReaction({
-          chatGuid: chatGuidForActions,
-          messageGuid: ackMessageId,
-          emoji: ackReactionValue,
-          remove: true,
-          opts: { cfg: config, accountId: account.accountId },
-        }).catch((err) => {
-          logVerbose(
-            core,
-            runtime,
-            `ack reaction removal failed chatGuid=${chatGuidForActions} msg=${ackMessageId}: ${String(err)}`,
-          );
-        });
+    if (sentMessage && chatGuidForActions && ackMessageId) {
+      core.channel.reactions.removeAckReactionAfterReply({
+        removeAfterReply: removeAckAfterReply,
+        ackReactionPromise,
+        ackReactionValue: ackReactionValue ?? null,
+        remove: () =>
+          sendBlueBubblesReaction({
+            chatGuid: chatGuidForActions,
+            messageGuid: ackMessageId,
+            emoji: ackReactionValue ?? "",
+            remove: true,
+            opts: { cfg: config, accountId: account.accountId },
+          }),
+        onError: (err) => {
+          logAckFailure({
+            log: (msg) => logVerbose(core, runtime, msg),
+            channel: "bluebubbles",
+            target: `${chatGuidForActions}/${ackMessageId}`,
+            error: err,
+          });
+        },
       });
     }
     if (chatGuidForActions && baseUrl && password && !sentMessage) {
@@ -1780,7 +1787,13 @@ async function processMessage(
         cfg: config,
         accountId: account.accountId,
       }).catch((err) => {
-        logVerbose(core, runtime, `typing stop (no reply) failed: ${String(err)}`);
+        logTypingFailure({
+          log: (msg) => logVerbose(core, runtime, msg),
+          channel: "bluebubbles",
+          action: "stop",
+          target: chatGuidForActions,
+          error: err,
+        });
       });
     }
   }

extensions/copilot-proxy/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clawdbot/copilot-proxy",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "Clawdbot Copilot Proxy provider plugin",
   "clawdbot": {

extensions/diagnostics-otel/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clawdbot/diagnostics-otel",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "Clawdbot diagnostics OpenTelemetry exporter",
   "clawdbot": {

extensions/discord/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clawdbot/discord",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "Clawdbot Discord channel plugin",
   "clawdbot": {

extensions/google-antigravity-auth/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clawdbot/google-antigravity-auth",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "Clawdbot Google Antigravity OAuth provider plugin",
   "clawdbot": {

extensions/google-gemini-cli-auth/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clawdbot/google-gemini-cli-auth",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "Clawdbot Gemini CLI OAuth provider plugin",
   "clawdbot": {

extensions/imessage/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clawdbot/imessage",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "Clawdbot iMessage channel plugin",
   "clawdbot": {

extensions/lobster/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clawdbot/lobster",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "Lobster workflow tool plugin (typed pipelines + resumable approvals)",
   "clawdbot": {

extensions/matrix/CHANGELOG.md

@@ -1,5 +1,10 @@
 # Changelog
 
+## 2026.1.23
+
+### Changes
+- Version alignment with core Clawdbot release numbers.
+
 ## 2026.1.22
 
 ### Changes

extensions/matrix/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clawdbot/matrix",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "Clawdbot Matrix channel plugin",
   "clawdbot": {

extensions/matrix/src/matrix/monitor/handler.ts

@@ -1,7 +1,12 @@
 import type { LocationMessageEventContent, MatrixClient } from "matrix-bot-sdk";
 
 import {
+  createReplyPrefixContext,
+  createTypingCallbacks,
   formatAllowlistMatchMeta,
+  logInboundDrop,
+  logTypingFailure,
+  resolveControlCommandGate,
   type RuntimeEnv,
 } from "clawdbot/plugin-sdk";
 import type { CoreConfig, ReplyToMode } from "../../types.js";
@@ -376,21 +381,25 @@ export function createMatrixRoomMessageHandler(params: MatrixMonitorHandlerParam
               userName: senderName,
             })
           : false;
-      const commandAuthorized = core.channel.commands.resolveCommandAuthorizedFromAuthorizers({
+      const hasControlCommandInMessage = core.channel.text.hasControlCommand(bodyText, cfg);
+      const commandGate = resolveControlCommandGate({
         useAccessGroups,
         authorizers: [
           { configured: effectiveAllowFrom.length > 0, allowed: senderAllowedForCommands },
           { configured: roomUsers.length > 0, allowed: senderAllowedForRoomUsers },
           { configured: groupAllowConfigured, allowed: senderAllowedForGroup },
         ],
+        allowTextCommands,
+        hasControlCommand: hasControlCommandInMessage,
       });
-      if (
-        isRoom &&
-        allowTextCommands &&
-        core.channel.text.hasControlCommand(bodyText, cfg) &&
-        !commandAuthorized
-      ) {
-        logVerboseMessage(`matrix: drop control command from unauthorized sender ${senderId}`);
+      const commandAuthorized = commandGate.commandAuthorized;
+      if (isRoom && commandGate.shouldBlock) {
+        logInboundDrop({
+          log: logVerboseMessage,
+          channel: "matrix",
+          reason: "control command (unauthorized)",
+          target: senderId,
+        });
         return;
       }
       const shouldRequireMention = isRoom
@@ -409,7 +418,8 @@ export function createMatrixRoomMessageHandler(params: MatrixMonitorHandlerParam
         !wasMentioned &&
         !hasExplicitMention &&
         commandAuthorized &&
-        core.channel.text.hasControlCommand(bodyText);
+        hasControlCommandInMessage;
+      const canDetectMention = mentionRegexes.length > 0 || hasExplicitMention;
       if (isRoom && shouldRequireMention && !wasMentioned && !shouldBypassMention) {
         logger.info({ roomId, reason: "no-mention" }, "skipping room message");
         return;
@@ -486,47 +496,45 @@ export function createMatrixRoomMessageHandler(params: MatrixMonitorHandlerParam
         OriginatingTo: `room:${roomId}`,
       });
 
-      void core.channel.session
-        .recordSessionMetaFromInbound({
-          storePath,
-          sessionKey: ctxPayload.SessionKey ?? route.sessionKey,
-          ctx: ctxPayload,
-        })
-        .catch((err) => {
+      await core.channel.session.recordInboundSession({
+        storePath,
+        sessionKey: ctxPayload.SessionKey ?? route.sessionKey,
+        ctx: ctxPayload,
+        updateLastRoute: isDirectMessage
+          ? {
+              sessionKey: route.mainSessionKey,
+              channel: "matrix",
+              to: `room:${roomId}`,
+              accountId: route.accountId,
+            }
+          : undefined,
+        onRecordError: (err) => {
           logger.warn(
             { error: String(err), storePath, sessionKey: ctxPayload.SessionKey ?? route.sessionKey },
             "failed updating session meta",
           );
-        });
-
-      if (isDirectMessage) {
-        await core.channel.session.updateLastRoute({
-          storePath,
-          sessionKey: route.mainSessionKey,
-          channel: "matrix",
-          to: `room:${roomId}`,
-          accountId: route.accountId,
-          ctx: ctxPayload,
-        });
-      }
+        },
+      });
 
       const preview = bodyText.slice(0, 200).replace(/\n/g, "\\n");
       logVerboseMessage(`matrix inbound: room=${roomId} from=${senderId} preview="${preview}"`);
 
       const ackReaction = (cfg.messages?.ackReaction ?? "").trim();
       const ackScope = cfg.messages?.ackReactionScope ?? "group-mentions";
-      const shouldAckReaction = () => {
-        if (!ackReaction) return false;
-        if (ackScope === "all") return true;
-        if (ackScope === "direct") return isDirectMessage;
-        if (ackScope === "group-all") return isRoom;
-        if (ackScope === "group-mentions") {
-          if (!isRoom) return false;
-          if (!shouldRequireMention) return false;
-          return wasMentioned || shouldBypassMention;
-        }
-        return false;
-      };
+      const shouldAckReaction = () =>
+        Boolean(
+          ackReaction &&
+            core.channel.reactions.shouldAckReaction({
+              scope: ackScope,
+              isDirect: isDirectMessage,
+              isGroup: isRoom,
+              isMentionableGroup: isRoom,
+              requireMention: Boolean(shouldRequireMention),
+              canDetectMention,
+              effectiveWasMentioned: wasMentioned || shouldBypassMention,
+              shouldBypassMention,
+            }),
+        );
       if (shouldAckReaction() && messageId) {
         reactMatrixMessage(roomId, messageId, ackReaction, client).catch((err) => {
           logVerboseMessage(`matrix react failed for room ${roomId}: ${String(err)}`);
@@ -553,10 +561,33 @@ export function createMatrixRoomMessageHandler(params: MatrixMonitorHandlerParam
         channel: "matrix",
         accountId: route.accountId,
       });
+      const prefixContext = createReplyPrefixContext({ cfg, agentId: route.agentId });
+      const typingCallbacks = createTypingCallbacks({
+        start: () => sendTypingMatrix(roomId, true, undefined, client),
+        stop: () => sendTypingMatrix(roomId, false, undefined, client),
+        onStartError: (err) => {
+          logTypingFailure({
+            log: logVerboseMessage,
+            channel: "matrix",
+            action: "start",
+            target: roomId,
+            error: err,
+          });
+        },
+        onStopError: (err) => {
+          logTypingFailure({
+            log: logVerboseMessage,
+            channel: "matrix",
+            action: "stop",
+            target: roomId,
+            error: err,
+          });
+        },
+      });
       const { dispatcher, replyOptions, markDispatchIdle } =
         core.channel.reply.createReplyDispatcherWithTyping({
-          responsePrefix: core.channel.reply.resolveEffectiveMessagesConfig(cfg, route.agentId)
-            .responsePrefix,
+          responsePrefix: prefixContext.responsePrefix,
+          responsePrefixContextProvider: prefixContext.responsePrefixContextProvider,
           humanDelay: core.channel.reply.resolveHumanDelayConfig(cfg, route.agentId),
           deliver: async (payload) => {
             await deliverMatrixReplies({
@@ -575,10 +606,8 @@ export function createMatrixRoomMessageHandler(params: MatrixMonitorHandlerParam
           onError: (err, info) => {
             runtime.error?.(`matrix ${info.kind} reply failed: ${String(err)}`);
           },
-          onReplyStart: () =>
-            sendTypingMatrix(roomId, true, undefined, client).catch(() => {}),
-          onIdle: () =>
-            sendTypingMatrix(roomId, false, undefined, client).catch(() => {}),
+          onReplyStart: typingCallbacks.onReplyStart,
+          onIdle: typingCallbacks.onIdle,
         });
 
       const { queuedFinal, counts } = await core.channel.reply.dispatchReplyFromConfig({
@@ -588,6 +617,7 @@ export function createMatrixRoomMessageHandler(params: MatrixMonitorHandlerParam
         replyOptions: {
           ...replyOptions,
           skillFilter: roomConfig?.skills,
+          onModelSelected: prefixContext.onModelSelected,
         },
       });
       markDispatchIdle();

extensions/mattermost/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clawdbot/mattermost",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "Clawdbot Mattermost channel plugin",
   "clawdbot": {

extensions/mattermost/src/mattermost/monitor.ts

@@ -7,10 +7,15 @@ import type {
   RuntimeEnv,
 } from "clawdbot/plugin-sdk";
 import {
+  createReplyPrefixContext,
+  createTypingCallbacks,
+  logInboundDrop,
+  logTypingFailure,
   buildPendingHistoryContextFromMap,
-  clearHistoryEntries,
+  clearHistoryEntriesIfEnabled,
   DEFAULT_GROUP_HISTORY_LIMIT,
-  recordPendingHistoryEntry,
+  recordPendingHistoryEntryIfEnabled,
+  resolveControlCommandGate,
   resolveChannelMediaMaxBytes,
   type HistoryEntry,
 } from "clawdbot/plugin-sdk";
@@ -30,12 +35,9 @@ import {
 } from "./client.js";
 import {
   createDedupeCache,
-  extractShortModelName,
   formatInboundFromLabel,
   rawDataToString,
-  resolveIdentityName,
   resolveThreadSessionKeys,
-  type ResponsePrefixContext,
 } from "./monitor-helpers.js";
 import { sendMessageMattermost } from "./send.js";
 
@@ -307,11 +309,7 @@ export async function monitorMattermostProvider(opts: MonitorMattermostOpts = {}
   };
 
   const sendTypingIndicator = async (channelId: string, parentId?: string) => {
-    try {
-      await sendMattermostTyping(client, { channelId, parentId });
-    } catch (err) {
-      logger.debug?.(`mattermost typing cue failed for channel ${channelId}: ${String(err)}`);
-    }
+    await sendMattermostTyping(client, { channelId, parentId });
   };
 
   const resolveChannelInfo = async (channelId: string): Promise<MattermostChannel | null> => {
@@ -403,7 +401,8 @@ export async function monitorMattermostProvider(opts: MonitorMattermostOpts = {}
       cfg,
       surface: "mattermost",
     });
-    const isControlCommand = allowTextCommands && core.channel.text.hasControlCommand(rawText, cfg);
+    const hasControlCommand = core.channel.text.hasControlCommand(rawText, cfg);
+    const isControlCommand = allowTextCommands && hasControlCommand;
     const useAccessGroups = cfg.commands?.useAccessGroups !== false;
     const senderAllowedForCommands = isSenderAllowed({
       senderId,
@@ -415,19 +414,20 @@ export async function monitorMattermostProvider(opts: MonitorMattermostOpts = {}
       senderName,
       allowFrom: effectiveGroupAllowFrom,
     });
+    const commandGate = resolveControlCommandGate({
+      useAccessGroups,
+      authorizers: [
+        { configured: effectiveAllowFrom.length > 0, allowed: senderAllowedForCommands },
+        {
+          configured: effectiveGroupAllowFrom.length > 0,
+          allowed: groupAllowedForCommands,
+        },
+      ],
+      allowTextCommands,
+      hasControlCommand,
+    });
     const commandAuthorized =
-      kind === "dm"
-        ? dmPolicy === "open" || senderAllowedForCommands
-        : core.channel.commands.resolveCommandAuthorizedFromAuthorizers({
-            useAccessGroups,
-            authorizers: [
-              { configured: effectiveAllowFrom.length > 0, allowed: senderAllowedForCommands },
-              {
-                configured: effectiveGroupAllowFrom.length > 0,
-                allowed: groupAllowedForCommands,
-              },
-            ],
-          });
+      kind === "dm" ? dmPolicy === "open" || senderAllowedForCommands : commandGate.commandAuthorized;
 
     if (kind === "dm") {
       if (dmPolicy === "disabled") {
@@ -488,10 +488,13 @@ export async function monitorMattermostProvider(opts: MonitorMattermostOpts = {}
       }
     }
 
-    if (kind !== "dm" && isControlCommand && !commandAuthorized) {
-      logVerboseMessage(
-        `mattermost: drop control command from unauthorized sender ${senderId}`,
-      );
+    if (kind !== "dm" && commandGate.shouldBlock) {
+      logInboundDrop({
+        log: logVerboseMessage,
+        channel: "mattermost",
+        reason: "control command (unauthorized)",
+        target: senderId,
+      });
       return;
     }
 
@@ -534,19 +537,19 @@ export async function monitorMattermostProvider(opts: MonitorMattermostOpts = {}
         : "");
     const pendingSender = senderName;
     const recordPendingHistory = () => {
-      if (!historyKey || historyLimit <= 0) return;
       const trimmed = pendingBody.trim();
-      if (!trimmed) return;
-      recordPendingHistoryEntry({
+      recordPendingHistoryEntryIfEnabled({
         historyMap: channelHistories,
-        historyKey,
         limit: historyLimit,
-        entry: {
-          sender: pendingSender,
-          body: trimmed,
-          timestamp: typeof post.create_at === "number" ? post.create_at : undefined,
-          messageId: post.id ?? undefined,
-        },
+        historyKey: historyKey ?? "",
+        entry: historyKey && trimmed
+          ? {
+              sender: pendingSender,
+              body: trimmed,
+              timestamp: typeof post.create_at === "number" ? post.create_at : undefined,
+              messageId: post.id ?? undefined,
+            }
+          : null,
       });
     };
 
@@ -623,7 +626,7 @@ export async function monitorMattermostProvider(opts: MonitorMattermostOpts = {}
       sender: { name: senderName, id: senderId },
     });
     let combinedBody = body;
-    if (historyKey && historyLimit > 0) {
+    if (historyKey) {
       combinedBody = buildPendingHistoryContextFromMap({
         historyMap: channelHistories,
         historyKey,
@@ -713,15 +716,23 @@ export async function monitorMattermostProvider(opts: MonitorMattermostOpts = {}
       accountId: account.accountId,
     });
 
-    let prefixContext: ResponsePrefixContext = {
-      identityName: resolveIdentityName(cfg, route.agentId),
-    };
+    const prefixContext = createReplyPrefixContext({ cfg, agentId: route.agentId });
 
+    const typingCallbacks = createTypingCallbacks({
+      start: () => sendTypingIndicator(channelId, threadRootId),
+      onStartError: (err) => {
+        logTypingFailure({
+          log: (message) => logger.debug?.(message),
+          channel: "mattermost",
+          target: channelId,
+          error: err,
+        });
+      },
+    });
     const { dispatcher, replyOptions, markDispatchIdle } =
       core.channel.reply.createReplyDispatcherWithTyping({
-        responsePrefix: core.channel.reply.resolveEffectiveMessagesConfig(cfg, route.agentId)
-          .responsePrefix,
-        responsePrefixContextProvider: () => prefixContext,
+        responsePrefix: prefixContext.responsePrefix,
+        responsePrefixContextProvider: prefixContext.responsePrefixContextProvider,
         humanDelay: core.channel.reply.resolveHumanDelayConfig(cfg, route.agentId),
         deliver: async (payload: ReplyPayload) => {
           const mediaUrls = payload.mediaUrls ?? (payload.mediaUrl ? [payload.mediaUrl] : []);
@@ -752,7 +763,7 @@ export async function monitorMattermostProvider(opts: MonitorMattermostOpts = {}
         onError: (err, info) => {
           runtime.error?.(`mattermost ${info.kind} reply failed: ${String(err)}`);
         },
-        onReplyStart: () => sendTypingIndicator(channelId, threadRootId),
+        onReplyStart: typingCallbacks.onReplyStart,
       });
 
     await core.channel.reply.dispatchReplyFromConfig({
@@ -763,17 +774,12 @@ export async function monitorMattermostProvider(opts: MonitorMattermostOpts = {}
         ...replyOptions,
         disableBlockStreaming:
           typeof account.blockStreaming === "boolean" ? !account.blockStreaming : undefined,
-        onModelSelected: (ctx) => {
-          prefixContext.provider = ctx.provider;
-          prefixContext.model = extractShortModelName(ctx.model);
-          prefixContext.modelFull = `${ctx.provider}/${ctx.model}`;
-          prefixContext.thinkingLevel = ctx.thinkLevel ?? "off";
-        },
+        onModelSelected: prefixContext.onModelSelected,
       },
     });
     markDispatchIdle();
-    if (historyKey && historyLimit > 0) {
-      clearHistoryEntries({ historyMap: channelHistories, historyKey });
+    if (historyKey) {
+      clearHistoryEntriesIfEnabled({ historyMap: channelHistories, historyKey, limit: historyLimit });
     }
   };
 

extensions/memory-core/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clawdbot/memory-core",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "Clawdbot core memory search plugin",
   "clawdbot": {

extensions/memory-lancedb/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clawdbot/memory-lancedb",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "Clawdbot LanceDB-backed long-term memory plugin with auto-recall/capture",
   "dependencies": {

extensions/msteams/CHANGELOG.md

@@ -1,5 +1,10 @@
 # Changelog
 
+## 2026.1.23
+
+### Changes
+- Version alignment with core Clawdbot release numbers.
+
 ## 2026.1.22
 
 ### Changes

extensions/msteams/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clawdbot/msteams",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "Clawdbot Microsoft Teams channel plugin",
   "clawdbot": {

extensions/msteams/src/attachments/download.ts

@@ -68,10 +68,10 @@ function scopeCandidatesForUrl(url: string): string[] {
       host.endsWith("1drv.ms") ||
       host.includes("sharepoint");
     return looksLikeGraph
-      ? ["https://graph.microsoft.com/.default", "https://api.botframework.com/.default"]
-      : ["https://api.botframework.com/.default", "https://graph.microsoft.com/.default"];
+      ? ["https://graph.microsoft.com", "https://api.botframework.com"]
+      : ["https://api.botframework.com", "https://graph.microsoft.com"];
   } catch {
-    return ["https://api.botframework.com/.default", "https://graph.microsoft.com/.default"];
+    return ["https://api.botframework.com", "https://graph.microsoft.com"];
   }
 }
 

extensions/msteams/src/attachments/graph.ts

@@ -198,7 +198,7 @@ export async function downloadMSTeamsGraphMedia(params: {
   const messageUrl = params.messageUrl;
   let accessToken: string;
   try {
-    accessToken = await params.tokenProvider.getAccessToken("https://graph.microsoft.com/.default");
+    accessToken = await params.tokenProvider.getAccessToken("https://graph.microsoft.com");
   } catch {
     return { media: [], messageUrl, tokenError: true };
   }

extensions/msteams/src/directory-live.ts

@@ -64,7 +64,7 @@ async function resolveGraphToken(cfg: unknown): Promise<string> {
   if (!creds) throw new Error("MS Teams credentials missing");
   const { sdk, authConfig } = await loadMSTeamsSdkWithAuth(creds);
   const tokenProvider = new sdk.MsalTokenProvider(authConfig);
-  const token = await tokenProvider.getAccessToken("https://graph.microsoft.com/.default");
+  const token = await tokenProvider.getAccessToken("https://graph.microsoft.com");
   const accessToken = readAccessToken(token);
   if (!accessToken) throw new Error("MS Teams graph token unavailable");
   return accessToken;

extensions/msteams/src/graph-upload.ts

@@ -13,7 +13,7 @@ import type { MSTeamsAccessTokenProvider } from "./attachments/types.js";
 
 const GRAPH_ROOT = "https://graph.microsoft.com/v1.0";
 const GRAPH_BETA = "https://graph.microsoft.com/beta";
-const GRAPH_SCOPE = "https://graph.microsoft.com/.default";
+const GRAPH_SCOPE = "https://graph.microsoft.com";
 
 export interface OneDriveUploadResult {
   id: string;

extensions/msteams/src/monitor-handler/message-handler.ts

@@ -1,8 +1,10 @@
 import {
   buildPendingHistoryContextFromMap,
-  clearHistoryEntries,
+  clearHistoryEntriesIfEnabled,
   DEFAULT_GROUP_HISTORY_LIMIT,
-  recordPendingHistoryEntry,
+  logInboundDrop,
+  recordPendingHistoryEntryIfEnabled,
+  resolveControlCommandGate,
   resolveMentionGating,
   formatAllowlistMatchMeta,
   type HistoryEntry,
@@ -251,15 +253,24 @@ export function createMSTeamsMessageHandler(deps: MSTeamsMessageHandlerDeps) {
       senderId,
       senderName,
     });
-    const commandAuthorized = core.channel.commands.resolveCommandAuthorizedFromAuthorizers({
+    const hasControlCommandInMessage = core.channel.text.hasControlCommand(text, cfg);
+    const commandGate = resolveControlCommandGate({
       useAccessGroups,
       authorizers: [
         { configured: effectiveDmAllowFrom.length > 0, allowed: ownerAllowedForCommands },
         { configured: effectiveGroupAllowFrom.length > 0, allowed: groupAllowedForCommands },
       ],
+      allowTextCommands: true,
+      hasControlCommand: hasControlCommandInMessage,
     });
-    if (core.channel.text.hasControlCommand(text, cfg) && !commandAuthorized) {
-      logVerboseMessage(`msteams: drop control command from unauthorized sender ${senderId}`);
+    const commandAuthorized = commandGate.commandAuthorized;
+    if (commandGate.shouldBlock) {
+      logInboundDrop({
+        log: logVerboseMessage,
+        channel: "msteams",
+        reason: "control command (unauthorized)",
+        target: senderId,
+      });
       return;
     }
 
@@ -371,19 +382,17 @@ export function createMSTeamsMessageHandler(deps: MSTeamsMessageHandlerDeps) {
           requireMention,
           mentioned,
         });
-        if (historyLimit > 0) {
-          recordPendingHistoryEntry({
-            historyMap: conversationHistories,
-            historyKey: conversationId,
-            limit: historyLimit,
-            entry: {
-              sender: senderName,
-              body: rawBody,
-              timestamp: timestamp?.getTime(),
-              messageId: activity.id ?? undefined,
-            },
-          });
-        }
+        recordPendingHistoryEntryIfEnabled({
+          historyMap: conversationHistories,
+          historyKey: conversationId,
+          limit: historyLimit,
+          entry: {
+            sender: senderName,
+            body: rawBody,
+            timestamp: timestamp?.getTime(),
+            messageId: activity.id ?? undefined,
+          },
+        });
         return;
       }
     }
@@ -426,7 +435,7 @@ export function createMSTeamsMessageHandler(deps: MSTeamsMessageHandlerDeps) {
     let combinedBody = body;
     const isRoomish = !isDirectMessage;
     const historyKey = isRoomish ? conversationId : undefined;
-    if (isRoomish && historyKey && historyLimit > 0) {
+    if (isRoomish && historyKey) {
       combinedBody = buildPendingHistoryContextFromMap({
         historyMap: conversationHistories,
         historyKey,
@@ -467,12 +476,13 @@ export function createMSTeamsMessageHandler(deps: MSTeamsMessageHandlerDeps) {
 	      ...mediaPayload,
 	    });
 
-	    void core.channel.session.recordSessionMetaFromInbound({
-	      storePath,
+    await core.channel.session.recordInboundSession({
+      storePath,
       sessionKey: ctxPayload.SessionKey ?? route.sessionKey,
       ctx: ctxPayload,
-    }).catch((err) => {
-      logVerboseMessage(`msteams: failed updating session meta: ${String(err)}`);
+      onRecordError: (err) => {
+        logVerboseMessage(`msteams: failed updating session meta: ${String(err)}`);
+      },
     });
 
     logVerboseMessage(`msteams inbound: from=${ctxPayload.From} preview="${preview}"`);
@@ -512,10 +522,11 @@ export function createMSTeamsMessageHandler(deps: MSTeamsMessageHandlerDeps) {
 
       const didSendReply = counts.final + counts.tool + counts.block > 0;
       if (!queuedFinal) {
-        if (isRoomish && historyKey && historyLimit > 0) {
-          clearHistoryEntries({
+        if (isRoomish && historyKey) {
+          clearHistoryEntriesIfEnabled({
             historyMap: conversationHistories,
             historyKey,
+            limit: historyLimit,
           });
         }
         return;
@@ -524,8 +535,12 @@ export function createMSTeamsMessageHandler(deps: MSTeamsMessageHandlerDeps) {
       logVerboseMessage(
         `msteams: delivered ${finalCount} reply${finalCount === 1 ? "" : "ies"} to ${teamsTo}`,
       );
-      if (isRoomish && historyKey && historyLimit > 0) {
-        clearHistoryEntries({ historyMap: conversationHistories, historyKey });
+      if (isRoomish && historyKey) {
+        clearHistoryEntriesIfEnabled({
+          historyMap: conversationHistories,
+          historyKey,
+          limit: historyLimit,
+        });
       }
     } catch (err) {
       log.error("dispatch failed", { error: String(err) });

extensions/msteams/src/reply-dispatcher.ts

@@ -1,4 +1,7 @@
 import {
+  createReplyPrefixContext,
+  createTypingCallbacks,
+  logTypingFailure,
   resolveChannelMediaMaxBytes,
   type ClawdbotConfig,
   type MSTeamsReplyStyle,
@@ -39,23 +42,33 @@ export function createMSTeamsReplyDispatcher(params: {
 }) {
   const core = getMSTeamsRuntime();
   const sendTypingIndicator = async () => {
-    try {
-      await params.context.sendActivities([{ type: "typing" }]);
-    } catch {
-      // Typing indicator is best-effort.
-    }
+    await params.context.sendActivities([{ type: "typing" }]);
   };
-
-  return core.channel.reply.createReplyDispatcherWithTyping({
-    responsePrefix: core.channel.reply.resolveEffectiveMessagesConfig(
-      params.cfg,
-      params.agentId,
-    ).responsePrefix,
-    humanDelay: core.channel.reply.resolveHumanDelayConfig(params.cfg, params.agentId),
-    deliver: async (payload) => {
-      const tableMode = core.channel.text.resolveMarkdownTableMode({
-        cfg: params.cfg,
+  const typingCallbacks = createTypingCallbacks({
+    start: sendTypingIndicator,
+    onStartError: (err) => {
+      logTypingFailure({
+        log: (message) => params.log.debug(message),
         channel: "msteams",
+        action: "start",
+        error: err,
+      });
+    },
+  });
+  const prefixContext = createReplyPrefixContext({
+    cfg: params.cfg,
+    agentId: params.agentId,
+  });
+
+  const { dispatcher, replyOptions, markDispatchIdle } =
+    core.channel.reply.createReplyDispatcherWithTyping({
+      responsePrefix: prefixContext.responsePrefix,
+      responsePrefixContextProvider: prefixContext.responsePrefixContextProvider,
+      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,
@@ -87,21 +100,27 @@ export function createMSTeamsReplyDispatcher(params: {
         mediaMaxBytes,
       });
       if (ids.length > 0) params.onSentMessageIds?.(ids);
-    },
-    onError: (err, info) => {
-      const errMsg = formatUnknownError(err);
-      const classification = classifyMSTeamsSendError(err);
-      const hint = formatMSTeamsSendErrorHint(classification);
-      params.runtime.error?.(
-        `msteams ${info.kind} reply failed: ${errMsg}${hint ? ` (${hint})` : ""}`,
-      );
-      params.log.error("reply failed", {
-        kind: info.kind,
-        error: errMsg,
-        classification,
-        hint,
-      });
-    },
-    onReplyStart: sendTypingIndicator,
-  });
+      },
+      onError: (err, info) => {
+        const errMsg = formatUnknownError(err);
+        const classification = classifyMSTeamsSendError(err);
+        const hint = formatMSTeamsSendErrorHint(classification);
+        params.runtime.error?.(
+          `msteams ${info.kind} reply failed: ${errMsg}${hint ? ` (${hint})` : ""}`,
+        );
+        params.log.error("reply failed", {
+          kind: info.kind,
+          error: errMsg,
+          classification,
+          hint,
+        });
+      },
+      onReplyStart: typingCallbacks.onReplyStart,
+    });
+
+  return {
+    dispatcher,
+    replyOptions: { ...replyOptions, onModelSelected: prefixContext.onModelSelected },
+    markDispatchIdle,
+  };
 }

extensions/msteams/src/resolve-allowlist.ts

@@ -143,7 +143,7 @@ async function resolveGraphToken(cfg: unknown): Promise<string> {
   if (!creds) throw new Error("MS Teams credentials missing");
   const { sdk, authConfig } = await loadMSTeamsSdkWithAuth(creds);
   const tokenProvider = new sdk.MsalTokenProvider(authConfig);
-  const token = await tokenProvider.getAccessToken("https://graph.microsoft.com/.default");
+  const token = await tokenProvider.getAccessToken("https://graph.microsoft.com");
   const accessToken = readAccessToken(token);
   if (!accessToken) throw new Error("MS Teams graph token unavailable");
   return accessToken;

extensions/nextcloud-talk/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clawdbot/nextcloud-talk",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "Clawdbot Nextcloud Talk channel plugin",
   "clawdbot": {

extensions/nextcloud-talk/src/inbound.ts

@@ -1,4 +1,9 @@
-import type { ClawdbotConfig, RuntimeEnv } from "clawdbot/plugin-sdk";
+import {
+  logInboundDrop,
+  resolveControlCommandGate,
+  type ClawdbotConfig,
+  type RuntimeEnv,
+} from "clawdbot/plugin-sdk";
 
 import type { ResolvedNextcloudTalkAccount } from "./accounts.js";
 import {
@@ -118,7 +123,11 @@ export async function handleNextcloudTalkInbound(params: {
     senderId,
     senderName,
   }).allowed;
-  const commandAuthorized = core.channel.commands.resolveCommandAuthorizedFromAuthorizers({
+  const hasControlCommand = core.channel.text.hasControlCommand(
+    rawBody,
+    config as ClawdbotConfig,
+  );
+  const commandGate = resolveControlCommandGate({
     useAccessGroups,
     authorizers: [
       {
@@ -127,7 +136,10 @@ export async function handleNextcloudTalkInbound(params: {
         allowed: senderAllowedForCommands,
       },
     ],
+    allowTextCommands,
+    hasControlCommand,
   });
+  const commandAuthorized = commandGate.commandAuthorized;
 
   if (isGroup) {
     const groupAllow = resolveNextcloudTalkGroupAllow({
@@ -188,15 +200,13 @@ export async function handleNextcloudTalkInbound(params: {
     }
   }
 
-  if (
-    isGroup &&
-    allowTextCommands &&
-    core.channel.text.hasControlCommand(rawBody, config as ClawdbotConfig) &&
-    commandAuthorized !== true
-  ) {
-    runtime.log?.(
-      `nextcloud-talk: drop control command from unauthorized sender ${senderId}`,
-    );
+  if (isGroup && commandGate.shouldBlock) {
+    logInboundDrop({
+      log: (message) => runtime.log?.(message),
+      channel: CHANNEL_ID,
+      reason: "control command (unauthorized)",
+      target: senderId,
+    });
     return;
   }
 
@@ -212,10 +222,6 @@ export async function handleNextcloudTalkInbound(params: {
         wildcardConfig: roomMatch.wildcardConfig,
       })
     : false;
-  const hasControlCommand = core.channel.text.hasControlCommand(
-    rawBody,
-    config as ClawdbotConfig,
-  );
   const mentionGate = resolveNextcloudTalkMentionGate({
     isGroup,
     requireMention: shouldRequireMention,
@@ -287,15 +293,14 @@ export async function handleNextcloudTalkInbound(params: {
     CommandAuthorized: commandAuthorized,
   });
 
-  void core.channel.session
-    .recordSessionMetaFromInbound({
-      storePath,
-      sessionKey: ctxPayload.SessionKey ?? route.sessionKey,
-      ctx: ctxPayload,
-    })
-    .catch((err) => {
+  await core.channel.session.recordInboundSession({
+    storePath,
+    sessionKey: ctxPayload.SessionKey ?? route.sessionKey,
+    ctx: ctxPayload,
+    onRecordError: (err) => {
       runtime.error?.(`nextcloud-talk: failed updating session meta: ${String(err)}`);
-    });
+    },
+  });
 
   await core.channel.reply.dispatchReplyWithBufferedBlockDispatcher({
     ctx: ctxPayload,

extensions/nostr/CHANGELOG.md

@@ -1,5 +1,10 @@
 # Changelog
 
+## 2026.1.23
+
+### Changes
+- Version alignment with core Clawdbot release numbers.
+
 ## 2026.1.22
 
 ### Changes

extensions/nostr/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clawdbot/nostr",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "Clawdbot Nostr channel plugin for NIP-04 encrypted DMs",
   "clawdbot": {

extensions/open-prose/package.json

@@ -1,9 +1,11 @@
 {
   "name": "@clawdbot/open-prose",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "OpenProse VM skill pack plugin (slash command + telemetry).",
   "clawdbot": {
-    "extensions": ["./index.ts"]
+    "extensions": [
+      "./index.ts"
+    ]
   }
 }

extensions/signal/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clawdbot/signal",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "Clawdbot Signal channel plugin",
   "clawdbot": {

extensions/slack/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clawdbot/slack",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "Clawdbot Slack channel plugin",
   "clawdbot": {

extensions/telegram/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clawdbot/telegram",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "Clawdbot Telegram channel plugin",
   "clawdbot": {

extensions/voice-call/CHANGELOG.md

@@ -1,5 +1,10 @@
 # Changelog
 
+## 2026.1.23
+
+### Changes
+- Version alignment with core Clawdbot release numbers.
+
 ## 2026.1.22
 
 ### Changes

extensions/voice-call/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clawdbot/voice-call",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "Clawdbot voice-call plugin",
   "dependencies": {

extensions/whatsapp/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clawdbot/whatsapp",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "Clawdbot WhatsApp channel plugin",
   "clawdbot": {

extensions/zalo/CHANGELOG.md

@@ -1,5 +1,10 @@
 # Changelog
 
+## 2026.1.23
+
+### Changes
+- Version alignment with core Clawdbot release numbers.
+
 ## 2026.1.22
 
 ### Changes

extensions/zalo/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clawdbot/zalo",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "Clawdbot Zalo channel plugin",
   "clawdbot": {

extensions/zalo/src/monitor.ts

@@ -570,12 +570,13 @@ async function processMessageWithPipeline(params: {
     OriginatingTo: `zalo:${chatId}`,
   });
 
-  void core.channel.session.recordSessionMetaFromInbound({
+  await core.channel.session.recordInboundSession({
     storePath,
     sessionKey: ctxPayload.SessionKey ?? route.sessionKey,
     ctx: ctxPayload,
-  }).catch((err) => {
-    runtime.error?.(`zalo: failed updating session meta: ${String(err)}`);
+    onRecordError: (err) => {
+      runtime.error?.(`zalo: failed updating session meta: ${String(err)}`);
+    },
   });
 
   const tableMode = core.channel.text.resolveMarkdownTableMode({

extensions/zalouser/CHANGELOG.md

@@ -1,5 +1,10 @@
 # Changelog
 
+## 2026.1.23
+
+### Changes
+- Version alignment with core Clawdbot release numbers.
+
 ## 2026.1.22
 
 ### Changes

extensions/zalouser/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clawdbot/zalouser",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "type": "module",
   "description": "Clawdbot Zalo Personal Account plugin via zca-cli",
   "dependencies": {

extensions/zalouser/src/monitor.ts

@@ -311,12 +311,13 @@ async function processMessage(
     OriginatingTo: `zalouser:${chatId}`,
   });
 
-  void core.channel.session.recordSessionMetaFromInbound({
+  await core.channel.session.recordInboundSession({
     storePath,
     sessionKey: ctxPayload.SessionKey ?? route.sessionKey,
     ctx: ctxPayload,
-  }).catch((err) => {
-    runtime.error?.(`zalouser: failed updating session meta: ${String(err)}`);
+    onRecordError: (err) => {
+      runtime.error?.(`zalouser: failed updating session meta: ${String(err)}`);
+    },
   });
 
   await core.channel.reply.dispatchReplyWithBufferedBlockDispatcher({

package.json

@@ -1,6 +1,6 @@
 {
   "name": "clawdbot",
-  "version": "2026.1.22",
+  "version": "2026.1.23",
   "description": "WhatsApp gateway CLI (Baileys web) with Pi RPC agent",
   "type": "module",
   "main": "dist/index.js",

pnpm-lock.yaml

@@ -376,8 +376,6 @@ importers:
         specifier: ^4.3.5
         version: 4.3.5
 
-  extensions/open-prose: {}
-
   extensions/signal: {}
 
   extensions/slack: {}

src/agents/compaction.test.ts

@@ -0,0 +1,107 @@
+import type { AgentMessage } from "@mariozechner/pi-agent-core";
+import { describe, expect, it } from "vitest";
+
+import {
+  estimateMessagesTokens,
+  pruneHistoryForContextShare,
+  splitMessagesByTokenShare,
+} from "./compaction.js";
+
+function makeMessage(id: number, size: number): AgentMessage {
+  return {
+    role: "user",
+    content: "x".repeat(size),
+    timestamp: id,
+  };
+}
+
+describe("splitMessagesByTokenShare", () => {
+  it("splits messages into two non-empty parts", () => {
+    const messages: AgentMessage[] = [
+      makeMessage(1, 4000),
+      makeMessage(2, 4000),
+      makeMessage(3, 4000),
+      makeMessage(4, 4000),
+    ];
+
+    const parts = splitMessagesByTokenShare(messages, 2);
+    expect(parts.length).toBeGreaterThanOrEqual(2);
+    expect(parts[0]?.length).toBeGreaterThan(0);
+    expect(parts[1]?.length).toBeGreaterThan(0);
+    expect(parts.flat().length).toBe(messages.length);
+  });
+
+  it("preserves message order across parts", () => {
+    const messages: AgentMessage[] = [
+      makeMessage(1, 4000),
+      makeMessage(2, 4000),
+      makeMessage(3, 4000),
+      makeMessage(4, 4000),
+      makeMessage(5, 4000),
+      makeMessage(6, 4000),
+    ];
+
+    const parts = splitMessagesByTokenShare(messages, 3);
+    expect(parts.flat().map((msg) => msg.timestamp)).toEqual(messages.map((msg) => msg.timestamp));
+  });
+});
+
+describe("pruneHistoryForContextShare", () => {
+  it("drops older chunks until the history budget is met", () => {
+    const messages: AgentMessage[] = [
+      makeMessage(1, 4000),
+      makeMessage(2, 4000),
+      makeMessage(3, 4000),
+      makeMessage(4, 4000),
+    ];
+    const maxContextTokens = 2000; // budget is 1000 tokens (50%)
+    const pruned = pruneHistoryForContextShare({
+      messages,
+      maxContextTokens,
+      maxHistoryShare: 0.5,
+      parts: 2,
+    });
+
+    expect(pruned.droppedChunks).toBeGreaterThan(0);
+    expect(pruned.keptTokens).toBeLessThanOrEqual(Math.floor(maxContextTokens * 0.5));
+    expect(pruned.messages.length).toBeGreaterThan(0);
+  });
+
+  it("keeps the newest messages when pruning", () => {
+    const messages: AgentMessage[] = [
+      makeMessage(1, 4000),
+      makeMessage(2, 4000),
+      makeMessage(3, 4000),
+      makeMessage(4, 4000),
+      makeMessage(5, 4000),
+      makeMessage(6, 4000),
+    ];
+    const totalTokens = estimateMessagesTokens(messages);
+    const maxContextTokens = Math.max(1, Math.floor(totalTokens * 0.5)); // budget = 25%
+    const pruned = pruneHistoryForContextShare({
+      messages,
+      maxContextTokens,
+      maxHistoryShare: 0.5,
+      parts: 2,
+    });
+
+    const keptIds = pruned.messages.map((msg) => msg.timestamp);
+    const expectedSuffix = messages.slice(-keptIds.length).map((msg) => msg.timestamp);
+    expect(keptIds).toEqual(expectedSuffix);
+  });
+
+  it("keeps history when already within budget", () => {
+    const messages: AgentMessage[] = [makeMessage(1, 1000)];
+    const maxContextTokens = 2000;
+    const pruned = pruneHistoryForContextShare({
+      messages,
+      maxContextTokens,
+      maxHistoryShare: 0.5,
+      parts: 2,
+    });
+
+    expect(pruned.droppedChunks).toBe(0);
+    expect(pruned.messages.length).toBe(messages.length);
+    expect(pruned.keptTokens).toBe(estimateMessagesTokens(messages));
+  });
+});

src/agents/compaction.ts

@@ -0,0 +1,341 @@
+import type { AgentMessage } from "@mariozechner/pi-agent-core";
+import type { ExtensionContext } from "@mariozechner/pi-coding-agent";
+import { estimateTokens, generateSummary } from "@mariozechner/pi-coding-agent";
+
+import { DEFAULT_CONTEXT_TOKENS } from "./defaults.js";
+
+export const BASE_CHUNK_RATIO = 0.4;
+export const MIN_CHUNK_RATIO = 0.15;
+export const SAFETY_MARGIN = 1.2; // 20% buffer for estimateTokens() inaccuracy
+const DEFAULT_SUMMARY_FALLBACK = "No prior history.";
+const DEFAULT_PARTS = 2;
+const MERGE_SUMMARIES_INSTRUCTIONS =
+  "Merge these partial summaries into a single cohesive summary. Preserve decisions," +
+  " TODOs, open questions, and any constraints.";
+
+export function estimateMessagesTokens(messages: AgentMessage[]): number {
+  return messages.reduce((sum, message) => sum + estimateTokens(message), 0);
+}
+
+function normalizeParts(parts: number, messageCount: number): number {
+  if (!Number.isFinite(parts) || parts <= 1) return 1;
+  return Math.min(Math.max(1, Math.floor(parts)), Math.max(1, messageCount));
+}
+
+export function splitMessagesByTokenShare(
+  messages: AgentMessage[],
+  parts = DEFAULT_PARTS,
+): AgentMessage[][] {
+  if (messages.length === 0) return [];
+  const normalizedParts = normalizeParts(parts, messages.length);
+  if (normalizedParts <= 1) return [messages];
+
+  const totalTokens = estimateMessagesTokens(messages);
+  const targetTokens = totalTokens / normalizedParts;
+  const chunks: AgentMessage[][] = [];
+  let current: AgentMessage[] = [];
+  let currentTokens = 0;
+
+  for (const message of messages) {
+    const messageTokens = estimateTokens(message);
+    if (
+      chunks.length < normalizedParts - 1 &&
+      current.length > 0 &&
+      currentTokens + messageTokens > targetTokens
+    ) {
+      chunks.push(current);
+      current = [];
+      currentTokens = 0;
+    }
+
+    current.push(message);
+    currentTokens += messageTokens;
+  }
+
+  if (current.length > 0) {
+    chunks.push(current);
+  }
+
+  return chunks;
+}
+
+export function chunkMessagesByMaxTokens(
+  messages: AgentMessage[],
+  maxTokens: number,
+): AgentMessage[][] {
+  if (messages.length === 0) return [];
+
+  const chunks: AgentMessage[][] = [];
+  let currentChunk: AgentMessage[] = [];
+  let currentTokens = 0;
+
+  for (const message of messages) {
+    const messageTokens = estimateTokens(message);
+    if (currentChunk.length > 0 && currentTokens + messageTokens > maxTokens) {
+      chunks.push(currentChunk);
+      currentChunk = [];
+      currentTokens = 0;
+    }
+
+    currentChunk.push(message);
+    currentTokens += messageTokens;
+
+    if (messageTokens > maxTokens) {
+      // Split oversized messages to avoid unbounded chunk growth.
+      chunks.push(currentChunk);
+      currentChunk = [];
+      currentTokens = 0;
+    }
+  }
+
+  if (currentChunk.length > 0) {
+    chunks.push(currentChunk);
+  }
+
+  return chunks;
+}
+
+/**
+ * Compute adaptive chunk ratio based on average message size.
+ * When messages are large, we use smaller chunks to avoid exceeding model limits.
+ */
+export function computeAdaptiveChunkRatio(messages: AgentMessage[], contextWindow: number): number {
+  if (messages.length === 0) return BASE_CHUNK_RATIO;
+
+  const totalTokens = estimateMessagesTokens(messages);
+  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.
+ */
+export 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"]>;
+  apiKey: string;
+  signal: AbortSignal;
+  reserveTokens: number;
+  maxChunkTokens: number;
+  customInstructions?: string;
+  previousSummary?: string;
+}): Promise<string> {
+  if (params.messages.length === 0) {
+    return params.previousSummary ?? DEFAULT_SUMMARY_FALLBACK;
+  }
+
+  const chunks = chunkMessagesByMaxTokens(params.messages, params.maxChunkTokens);
+  let summary = params.previousSummary;
+
+  for (const chunk of chunks) {
+    summary = await generateSummary(
+      chunk,
+      params.model,
+      params.reserveTokens,
+      params.apiKey,
+      params.signal,
+      params.customInstructions,
+      summary,
+    );
+  }
+
+  return summary ?? DEFAULT_SUMMARY_FALLBACK;
+}
+
+/**
+ * Summarize with progressive fallback for handling oversized messages.
+ * If full summarization fails, tries partial summarization excluding oversized messages.
+ */
+export 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 ?? DEFAULT_SUMMARY_FALLBACK;
+  }
+
+  // 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 async function summarizeInStages(params: {
+  messages: AgentMessage[];
+  model: NonNullable<ExtensionContext["model"]>;
+  apiKey: string;
+  signal: AbortSignal;
+  reserveTokens: number;
+  maxChunkTokens: number;
+  contextWindow: number;
+  customInstructions?: string;
+  previousSummary?: string;
+  parts?: number;
+  minMessagesForSplit?: number;
+}): Promise<string> {
+  const { messages } = params;
+  if (messages.length === 0) {
+    return params.previousSummary ?? DEFAULT_SUMMARY_FALLBACK;
+  }
+
+  const minMessagesForSplit = Math.max(2, params.minMessagesForSplit ?? 4);
+  const parts = normalizeParts(params.parts ?? DEFAULT_PARTS, messages.length);
+  const totalTokens = estimateMessagesTokens(messages);
+
+  if (parts <= 1 || messages.length < minMessagesForSplit || totalTokens <= params.maxChunkTokens) {
+    return summarizeWithFallback(params);
+  }
+
+  const splits = splitMessagesByTokenShare(messages, parts).filter((chunk) => chunk.length > 0);
+  if (splits.length <= 1) {
+    return summarizeWithFallback(params);
+  }
+
+  const partialSummaries: string[] = [];
+  for (const chunk of splits) {
+    partialSummaries.push(
+      await summarizeWithFallback({
+        ...params,
+        messages: chunk,
+        previousSummary: undefined,
+      }),
+    );
+  }
+
+  if (partialSummaries.length === 1) {
+    return partialSummaries[0];
+  }
+
+  const summaryMessages: AgentMessage[] = partialSummaries.map((summary) => ({
+    role: "user",
+    content: summary,
+    timestamp: Date.now(),
+  }));
+
+  const mergeInstructions = params.customInstructions
+    ? `${MERGE_SUMMARIES_INSTRUCTIONS}\n\nAdditional focus:\n${params.customInstructions}`
+    : MERGE_SUMMARIES_INSTRUCTIONS;
+
+  return summarizeWithFallback({
+    ...params,
+    messages: summaryMessages,
+    customInstructions: mergeInstructions,
+  });
+}
+
+export function pruneHistoryForContextShare(params: {
+  messages: AgentMessage[];
+  maxContextTokens: number;
+  maxHistoryShare?: number;
+  parts?: number;
+}): {
+  messages: AgentMessage[];
+  droppedChunks: number;
+  droppedMessages: number;
+  droppedTokens: number;
+  keptTokens: number;
+  budgetTokens: number;
+} {
+  const maxHistoryShare = params.maxHistoryShare ?? 0.5;
+  const budgetTokens = Math.max(1, Math.floor(params.maxContextTokens * maxHistoryShare));
+  let keptMessages = params.messages;
+  let droppedChunks = 0;
+  let droppedMessages = 0;
+  let droppedTokens = 0;
+
+  const parts = normalizeParts(params.parts ?? DEFAULT_PARTS, keptMessages.length);
+
+  while (keptMessages.length > 0 && estimateMessagesTokens(keptMessages) > budgetTokens) {
+    const chunks = splitMessagesByTokenShare(keptMessages, parts);
+    if (chunks.length <= 1) break;
+    const [dropped, ...rest] = chunks;
+    droppedChunks += 1;
+    droppedMessages += dropped.length;
+    droppedTokens += estimateMessagesTokens(dropped);
+    keptMessages = rest.flat();
+  }
+
+  return {
+    messages: keptMessages,
+    droppedChunks,
+    droppedMessages,
+    droppedTokens,
+    keptTokens: estimateMessagesTokens(keptMessages),
+    budgetTokens,
+  };
+}
+
+export function resolveContextWindowTokens(model?: ExtensionContext["model"]): number {
+  return Math.max(1, Math.floor(model?.contextWindow ?? DEFAULT_CONTEXT_TOKENS));
+}

src/agents/pi-embedded-runner/run.ts

@@ -79,6 +79,7 @@ export async function runEmbeddedPiAgent(
         ? "markdown"
         : "plain"
       : "markdown");
+  const isProbeSession = params.sessionId?.startsWith("probe-") ?? false;
 
   return enqueueCommandInLane(sessionLane, () =>
     enqueueGlobal(async () => {
@@ -455,7 +456,7 @@ export async function runEmbeddedPiAgent(
                 cfg: params.config,
                 agentDir: params.agentDir,
               });
-              if (timedOut) {
+              if (timedOut && !isProbeSession) {
                 log.warn(
                   `Profile ${lastProfileId} timed out (possible rate limit). Trying next account...`,
                 );

src/agents/pi-embedded-runner/run/attempt.ts

@@ -595,18 +595,23 @@ export async function runEmbeddedAttempt(
       setActiveEmbeddedRun(params.sessionId, queueHandle);
 
       let abortWarnTimer: NodeJS.Timeout | undefined;
+      const isProbeSession = params.sessionId?.startsWith("probe-") ?? false;
       const abortTimer = setTimeout(
         () => {
-          log.warn(
-            `embedded run timeout: runId=${params.runId} sessionId=${params.sessionId} timeoutMs=${params.timeoutMs}`,
-          );
+          if (!isProbeSession) {
+            log.warn(
+              `embedded run timeout: runId=${params.runId} sessionId=${params.sessionId} timeoutMs=${params.timeoutMs}`,
+            );
+          }
           abortRun(true);
           if (!abortWarnTimer) {
             abortWarnTimer = setTimeout(() => {
               if (!activeSession.isStreaming) return;
-              log.warn(
-                `embedded run abort still streaming: runId=${params.runId} sessionId=${params.sessionId}`,
-              );
+              if (!isProbeSession) {
+                log.warn(
+                  `embedded run abort still streaming: runId=${params.runId} sessionId=${params.sessionId}`,
+                );
+              }
             }, 10_000);
           }
         },

src/agents/pi-extensions/compaction-safeguard.ts

@@ -1,12 +1,16 @@
 import type { AgentMessage } from "@mariozechner/pi-agent-core";
-import type { ExtensionAPI, ExtensionContext, FileOperations } from "@mariozechner/pi-coding-agent";
-import { estimateTokens, generateSummary } from "@mariozechner/pi-coding-agent";
-
-import { DEFAULT_CONTEXT_TOKENS } from "../defaults.js";
-
-const BASE_CHUNK_RATIO = 0.4;
-const MIN_CHUNK_RATIO = 0.15;
-const SAFETY_MARGIN = 1.2; // 20% buffer for estimateTokens() inaccuracy
+import type { ExtensionAPI, FileOperations } from "@mariozechner/pi-coding-agent";
+import {
+  BASE_CHUNK_RATIO,
+  MIN_CHUNK_RATIO,
+  SAFETY_MARGIN,
+  computeAdaptiveChunkRatio,
+  estimateMessagesTokens,
+  isOversizedForSummary,
+  pruneHistoryForContextShare,
+  resolveContextWindowTokens,
+  summarizeInStages,
+} from "../compaction.js";
 const FALLBACK_SUMMARY =
   "Summary unavailable due to context limits. Older messages were truncated.";
 const TURN_PREFIX_INSTRUCTIONS =
@@ -129,175 +133,6 @@ function formatFileOperations(readFiles: string[], modifiedFiles: string[]): str
   return `\n\n${sections.join("\n\n")}`;
 }
 
-function chunkMessages(messages: AgentMessage[], maxTokens: number): AgentMessage[][] {
-  if (messages.length === 0) return [];
-
-  const chunks: AgentMessage[][] = [];
-  let currentChunk: AgentMessage[] = [];
-  let currentTokens = 0;
-
-  for (const message of messages) {
-    const messageTokens = estimateTokens(message);
-    if (currentChunk.length > 0 && currentTokens + messageTokens > maxTokens) {
-      chunks.push(currentChunk);
-      currentChunk = [];
-      currentTokens = 0;
-    }
-
-    currentChunk.push(message);
-    currentTokens += messageTokens;
-
-    if (messageTokens > maxTokens) {
-      // Split oversized messages to avoid unbounded chunk growth.
-      chunks.push(currentChunk);
-      currentChunk = [];
-      currentTokens = 0;
-    }
-  }
-
-  if (currentChunk.length > 0) {
-    chunks.push(currentChunk);
-  }
-
-  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"]>;
-  apiKey: string;
-  signal: AbortSignal;
-  reserveTokens: number;
-  maxChunkTokens: number;
-  customInstructions?: string;
-  previousSummary?: string;
-}): Promise<string> {
-  if (params.messages.length === 0) {
-    return params.previousSummary ?? "No prior history.";
-  }
-
-  const chunks = chunkMessages(params.messages, params.maxChunkTokens);
-  let summary = params.previousSummary;
-
-  for (const chunk of chunks) {
-    summary = await generateSummary(
-      chunk,
-      params.model,
-      params.reserveTokens,
-      params.apiKey,
-      params.signal,
-      params.customInstructions,
-      summary,
-    );
-  }
-
-  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;
@@ -335,19 +170,48 @@ export default function compactionSafeguardExtension(api: ExtensionAPI): void {
     }
 
     try {
-      const contextWindowTokens = Math.max(
-        1,
-        Math.floor(model.contextWindow ?? DEFAULT_CONTEXT_TOKENS),
-      );
+      const contextWindowTokens = resolveContextWindowTokens(model);
+      const turnPrefixMessages = preparation.turnPrefixMessages ?? [];
+      let messagesToSummarize = preparation.messagesToSummarize;
+
+      const tokensBefore =
+        typeof preparation.tokensBefore === "number" && Number.isFinite(preparation.tokensBefore)
+          ? preparation.tokensBefore
+          : undefined;
+      if (tokensBefore !== undefined) {
+        const summarizableTokens =
+          estimateMessagesTokens(messagesToSummarize) + estimateMessagesTokens(turnPrefixMessages);
+        const newContentTokens = Math.max(0, Math.floor(tokensBefore - summarizableTokens));
+        const maxHistoryTokens = Math.floor(contextWindowTokens * 0.5);
+
+        if (newContentTokens > maxHistoryTokens) {
+          const pruned = pruneHistoryForContextShare({
+            messages: messagesToSummarize,
+            maxContextTokens: contextWindowTokens,
+            maxHistoryShare: 0.5,
+            parts: 2,
+          });
+          if (pruned.droppedChunks > 0) {
+            const newContentRatio = (newContentTokens / contextWindowTokens) * 100;
+            console.warn(
+              `Compaction safeguard: new content uses ${newContentRatio.toFixed(
+                1,
+              )}% of context; dropped ${pruned.droppedChunks} older chunk(s) ` +
+                `(${pruned.droppedMessages} messages) to fit history budget.`,
+            );
+            messagesToSummarize = pruned.messages;
+          }
+        }
+      }
 
       // Use adaptive chunk ratio based on message sizes
-      const allMessages = [...preparation.messagesToSummarize, ...preparation.turnPrefixMessages];
+      const allMessages = [...messagesToSummarize, ...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 summarizeWithFallback({
-        messages: preparation.messagesToSummarize,
+      const historySummary = await summarizeInStages({
+        messages: messagesToSummarize,
         model,
         apiKey,
         signal,
@@ -359,9 +223,9 @@ export default function compactionSafeguardExtension(api: ExtensionAPI): void {
       });
 
       let summary = historySummary;
-      if (preparation.isSplitTurn && preparation.turnPrefixMessages.length > 0) {
-        const prefixSummary = await summarizeWithFallback({
-          messages: preparation.turnPrefixMessages,
+      if (preparation.isSplitTurn && turnPrefixMessages.length > 0) {
+        const prefixSummary = await summarizeInStages({
+          messages: turnPrefixMessages,
           model,
           apiKey,
           signal,
@@ -369,6 +233,7 @@ export default function compactionSafeguardExtension(api: ExtensionAPI): void {
           maxChunkTokens,
           contextWindow: contextWindowTokens,
           customInstructions: TURN_PREFIX_INSTRUCTIONS,
+          previousSummary: undefined,
         });
         summary = `${historySummary}\n\n---\n\n**Turn Context (split turn):**\n\n${prefixSummary}`;
       }

src/agents/sandbox-docker-image-check-errors.test.ts

@@ -0,0 +1,19 @@
+import { describe, expect, it } from "vitest";
+
+import { resolveDockerImageInspectResult } from "./sandbox/docker.js";
+
+describe("ensureDockerImage", () => {
+  it("surfaces inspect failures with detail", async () => {
+    const result = resolveDockerImageInspectResult("custom-sandbox:latest", {
+      stdout: "",
+      stderr: "permission denied while trying to connect to the Docker daemon socket",
+      code: 1,
+    });
+
+    expect(result).toEqual({
+      exists: false,
+      error:
+        "Failed to inspect sandbox image: permission denied while trying to connect to the Docker daemon socket",
+    });
+  });
+});

src/agents/sandbox/docker.ts

@@ -46,11 +46,25 @@ export async function readDockerPort(containerName: string, port: number) {
   return Number.isFinite(mapped) ? mapped : null;
 }
 
+export function resolveDockerImageInspectResult(
+  image: string,
+  result: { stdout: string; stderr: string; code: number },
+): { exists: boolean; error?: string } {
+  if (result.code === 0) return { exists: true };
+  const stderr = result.stderr.trim();
+  if (/no such image/i.test(stderr)) return { exists: false };
+  const stdout = result.stdout.trim();
+  const detail = stderr || stdout || `docker image inspect ${image} failed (exit ${result.code})`;
+  return { exists: false, error: `Failed to inspect sandbox image: ${detail}` };
+}
+
 async function dockerImageExists(image: string) {
   const result = await execDocker(["image", "inspect", image], {
     allowFailure: true,
   });
-  return result.code === 0;
+  const resolved = resolveDockerImageInspectResult(image, result);
+  if (resolved.error) throw new Error(resolved.error);
+  return resolved.exists;
 }
 
 export async function ensureDockerImage(image: string) {

src/auto-reply/dispatch.ts

@@ -0,0 +1,77 @@
+import type { ClawdbotConfig } from "../config/config.js";
+import type { FinalizedMsgContext, MsgContext } from "./templating.js";
+import type { GetReplyOptions } from "./types.js";
+import { finalizeInboundContext } from "./reply/inbound-context.js";
+import type { DispatchFromConfigResult } from "./reply/dispatch-from-config.js";
+import { dispatchReplyFromConfig } from "./reply/dispatch-from-config.js";
+import {
+  createReplyDispatcher,
+  createReplyDispatcherWithTyping,
+  type ReplyDispatcher,
+  type ReplyDispatcherOptions,
+  type ReplyDispatcherWithTypingOptions,
+} from "./reply/reply-dispatcher.js";
+
+export type DispatchInboundResult = DispatchFromConfigResult;
+
+export async function dispatchInboundMessage(params: {
+  ctx: MsgContext | FinalizedMsgContext;
+  cfg: ClawdbotConfig;
+  dispatcher: ReplyDispatcher;
+  replyOptions?: Omit<GetReplyOptions, "onToolResult" | "onBlockReply">;
+  replyResolver?: typeof import("./reply.js").getReplyFromConfig;
+}): Promise<DispatchInboundResult> {
+  const finalized = finalizeInboundContext(params.ctx);
+  return await dispatchReplyFromConfig({
+    ctx: finalized,
+    cfg: params.cfg,
+    dispatcher: params.dispatcher,
+    replyOptions: params.replyOptions,
+    replyResolver: params.replyResolver,
+  });
+}
+
+export async function dispatchInboundMessageWithBufferedDispatcher(params: {
+  ctx: MsgContext | FinalizedMsgContext;
+  cfg: ClawdbotConfig;
+  dispatcherOptions: ReplyDispatcherWithTypingOptions;
+  replyOptions?: Omit<GetReplyOptions, "onToolResult" | "onBlockReply">;
+  replyResolver?: typeof import("./reply.js").getReplyFromConfig;
+}): Promise<DispatchInboundResult> {
+  const { dispatcher, replyOptions, markDispatchIdle } = createReplyDispatcherWithTyping(
+    params.dispatcherOptions,
+  );
+
+  const result = await dispatchInboundMessage({
+    ctx: params.ctx,
+    cfg: params.cfg,
+    dispatcher,
+    replyResolver: params.replyResolver,
+    replyOptions: {
+      ...params.replyOptions,
+      ...replyOptions,
+    },
+  });
+
+  markDispatchIdle();
+  return result;
+}
+
+export async function dispatchInboundMessageWithDispatcher(params: {
+  ctx: MsgContext | FinalizedMsgContext;
+  cfg: ClawdbotConfig;
+  dispatcherOptions: ReplyDispatcherOptions;
+  replyOptions?: Omit<GetReplyOptions, "onToolResult" | "onBlockReply">;
+  replyResolver?: typeof import("./reply.js").getReplyFromConfig;
+}): Promise<DispatchInboundResult> {
+  const dispatcher = createReplyDispatcher(params.dispatcherOptions);
+  const result = await dispatchInboundMessage({
+    ctx: params.ctx,
+    cfg: params.cfg,
+    dispatcher,
+    replyResolver: params.replyResolver,
+    replyOptions: params.replyOptions,
+  });
+  await dispatcher.waitForIdle();
+  return result;
+}

src/auto-reply/heartbeat.test.ts

@@ -1,10 +1,6 @@
 import { describe, expect, it } from "vitest";
 
-import {
-  DEFAULT_HEARTBEAT_ACK_MAX_CHARS,
-  isHeartbeatContentEffectivelyEmpty,
-  stripHeartbeatToken,
-} from "./heartbeat.js";
+import { DEFAULT_HEARTBEAT_ACK_MAX_CHARS, stripHeartbeatToken } from "./heartbeat.js";
 import { HEARTBEAT_TOKEN } from "./tokens.js";
 
 describe("stripHeartbeatToken", () => {
@@ -109,76 +105,3 @@ describe("stripHeartbeatToken", () => {
     });
   });
 });
-
-describe("isHeartbeatContentEffectivelyEmpty", () => {
-  it("returns false for undefined/null (missing file should not skip)", () => {
-    expect(isHeartbeatContentEffectivelyEmpty(undefined)).toBe(false);
-    expect(isHeartbeatContentEffectivelyEmpty(null)).toBe(false);
-  });
-
-  it("returns true for empty string", () => {
-    expect(isHeartbeatContentEffectivelyEmpty("")).toBe(true);
-  });
-
-  it("returns true for whitespace only", () => {
-    expect(isHeartbeatContentEffectivelyEmpty("   ")).toBe(true);
-    expect(isHeartbeatContentEffectivelyEmpty("\n\n\n")).toBe(true);
-    expect(isHeartbeatContentEffectivelyEmpty("  \n  \n  ")).toBe(true);
-    expect(isHeartbeatContentEffectivelyEmpty("\t\t")).toBe(true);
-  });
-
-  it("returns true for header-only content", () => {
-    expect(isHeartbeatContentEffectivelyEmpty("# HEARTBEAT.md")).toBe(true);
-    expect(isHeartbeatContentEffectivelyEmpty("# HEARTBEAT.md\n")).toBe(true);
-    expect(isHeartbeatContentEffectivelyEmpty("# HEARTBEAT.md\n\n")).toBe(true);
-  });
-
-  it("returns true for comments only", () => {
-    expect(isHeartbeatContentEffectivelyEmpty("# Header\n# Another comment")).toBe(true);
-    expect(isHeartbeatContentEffectivelyEmpty("## Subheader\n### Another")).toBe(true);
-  });
-
-  it("returns true for default template content (header + comment)", () => {
-    const defaultTemplate = `# HEARTBEAT.md
-
-Keep this file empty unless you want a tiny checklist. Keep it small.
-`;
-    // Note: The template has actual text content, so it's NOT effectively empty
-    expect(isHeartbeatContentEffectivelyEmpty(defaultTemplate)).toBe(false);
-  });
-
-  it("returns true for header with only empty lines", () => {
-    expect(isHeartbeatContentEffectivelyEmpty("# HEARTBEAT.md\n\n\n")).toBe(true);
-  });
-
-  it("returns false when actionable content exists", () => {
-    expect(isHeartbeatContentEffectivelyEmpty("- Check email")).toBe(false);
-    expect(isHeartbeatContentEffectivelyEmpty("# HEARTBEAT.md\n- Task 1")).toBe(false);
-    expect(isHeartbeatContentEffectivelyEmpty("Remind me to call mom")).toBe(false);
-  });
-
-  it("returns false for content with tasks after header", () => {
-    const content = `# HEARTBEAT.md
-
-- Task 1
-- Task 2
-`;
-    expect(isHeartbeatContentEffectivelyEmpty(content)).toBe(false);
-  });
-
-  it("returns false for mixed content with non-comment text", () => {
-    const content = `# HEARTBEAT.md
-## Tasks
-Check the server logs
-`;
-    expect(isHeartbeatContentEffectivelyEmpty(content)).toBe(false);
-  });
-
-  it("treats markdown headers as comments (effectively empty)", () => {
-    const content = `# HEARTBEAT.md
-## Section 1
-### Subsection
-`;
-    expect(isHeartbeatContentEffectivelyEmpty(content)).toBe(true);
-  });
-});

src/auto-reply/heartbeat.ts

@@ -7,38 +7,6 @@ export const HEARTBEAT_PROMPT =
 export const DEFAULT_HEARTBEAT_EVERY = "30m";
 export const DEFAULT_HEARTBEAT_ACK_MAX_CHARS = 300;
 
-/**
- * Check if HEARTBEAT.md content is "effectively empty" - meaning it has no actionable tasks.
- * This allows skipping heartbeat API calls when no tasks are configured.
- *
- * A file is considered effectively empty if it contains only:
- * - Whitespace
- * - Comment lines (lines starting with #)
- * - Empty lines
- *
- * Note: A missing file returns false (not effectively empty) so the LLM can still
- * decide what to do. This function is only for when the file exists but has no content.
- */
-export function isHeartbeatContentEffectivelyEmpty(content: string | undefined | null): boolean {
-  if (content === undefined || content === null) return false;
-  if (typeof content !== "string") return false;
-
-  const lines = content.split("\n");
-  for (const line of lines) {
-    const trimmed = line.trim();
-    // Skip empty lines
-    if (!trimmed) continue;
-    // Skip markdown header lines (# followed by space or EOL, ## etc)
-    // This intentionally does NOT skip lines like "#TODO" or "#hashtag" which might be content
-    // (Those aren't valid markdown headers - ATX headers require space after #)
-    if (/^#+(\s|$)/.test(trimmed)) continue;
-    // Found a non-empty, non-comment line - there's actionable content
-    return false;
-  }
-  // All lines were either empty or comments
-  return true;
-}
-
 export function resolveHeartbeatPrompt(raw?: string): string {
   const trimmed = typeof raw === "string" ? raw.trim() : "";
   return trimmed || HEARTBEAT_PROMPT;

src/auto-reply/reply/agent-runner-execution.ts

@@ -82,7 +82,8 @@ export async function runAgentTurnWithFallback(params: {
   // Track payloads sent directly (not via pipeline) during tool flush to avoid duplicates.
   const directlySentBlockKeys = new Set<string>();
 
-  const runId = crypto.randomUUID();
+  const runId = params.opts?.runId ?? crypto.randomUUID();
+  params.opts?.onAgentRunStart?.(runId);
   if (params.sessionKey) {
     registerAgentRunContext(runId, {
       sessionKey: params.sessionKey,
@@ -174,6 +175,7 @@ export async function runAgentTurnWithFallback(params: {
               extraSystemPrompt: params.followupRun.run.extraSystemPrompt,
               ownerNumbers: params.followupRun.run.ownerNumbers,
               cliSessionId,
+              images: params.opts?.images,
             })
               .then((result) => {
                 emitAgentEvent({
@@ -248,6 +250,8 @@ export async function runAgentTurnWithFallback(params: {
             bashElevated: params.followupRun.run.bashElevated,
             timeoutMs: params.followupRun.run.timeoutMs,
             runId,
+            images: params.opts?.images,
+            abortSignal: params.opts?.abortSignal,
             blockReplyBreak: params.resolvedBlockStreamingBreak,
             blockReplyChunking: params.blockReplyChunking,
             onPartialReply: allowPartialStream

src/auto-reply/reply/history.test.ts

@@ -5,7 +5,9 @@ import {
   buildHistoryContextFromEntries,
   buildHistoryContextFromMap,
   buildPendingHistoryContextFromMap,
+  clearHistoryEntriesIfEnabled,
   HISTORY_CONTEXT_MARKER,
+  recordPendingHistoryEntryIfEnabled,
 } from "./history.js";
 import { CURRENT_MESSAGE_MARKER } from "./mentions.js";
 
@@ -105,4 +107,46 @@ describe("history helpers", () => {
     expect(result).toContain(CURRENT_MESSAGE_MARKER);
     expect(result).toContain("current");
   });
+
+  it("records pending entries only when enabled", () => {
+    const historyMap = new Map<string, { sender: string; body: string }[]>();
+
+    recordPendingHistoryEntryIfEnabled({
+      historyMap,
+      historyKey: "group",
+      limit: 0,
+      entry: { sender: "A", body: "one" },
+    });
+    expect(historyMap.get("group")).toEqual(undefined);
+
+    recordPendingHistoryEntryIfEnabled({
+      historyMap,
+      historyKey: "group",
+      limit: 2,
+      entry: null,
+    });
+    expect(historyMap.get("group")).toEqual(undefined);
+
+    recordPendingHistoryEntryIfEnabled({
+      historyMap,
+      historyKey: "group",
+      limit: 2,
+      entry: { sender: "B", body: "two" },
+    });
+    expect(historyMap.get("group")?.map((entry) => entry.body)).toEqual(["two"]);
+  });
+
+  it("clears history entries only when enabled", () => {
+    const historyMap = new Map<string, { sender: string; body: string }[]>();
+    historyMap.set("group", [
+      { sender: "A", body: "one" },
+      { sender: "B", body: "two" },
+    ]);
+
+    clearHistoryEntriesIfEnabled({ historyMap, historyKey: "group", limit: 0 });
+    expect(historyMap.get("group")?.map((entry) => entry.body)).toEqual(["one", "two"]);
+
+    clearHistoryEntriesIfEnabled({ historyMap, historyKey: "group", limit: 2 });
+    expect(historyMap.get("group")).toEqual([]);
+  });
 });

src/auto-reply/reply/history.ts

@@ -47,6 +47,22 @@ export function recordPendingHistoryEntry<T extends HistoryEntry>(params: {
   return appendHistoryEntry(params);
 }
 
+export function recordPendingHistoryEntryIfEnabled<T extends HistoryEntry>(params: {
+  historyMap: Map<string, T[]>;
+  historyKey: string;
+  entry?: T | null;
+  limit: number;
+}): T[] {
+  if (!params.entry) return [];
+  if (params.limit <= 0) return [];
+  return recordPendingHistoryEntry({
+    historyMap: params.historyMap,
+    historyKey: params.historyKey,
+    entry: params.entry,
+    limit: params.limit,
+  });
+}
+
 export function buildPendingHistoryContextFromMap(params: {
   historyMap: Map<string, HistoryEntry[]>;
   historyKey: string;
@@ -101,6 +117,15 @@ export function clearHistoryEntries(params: {
   params.historyMap.set(params.historyKey, []);
 }
 
+export function clearHistoryEntriesIfEnabled(params: {
+  historyMap: Map<string, HistoryEntry[]>;
+  historyKey: string;
+  limit: number;
+}): void {
+  if (params.limit <= 0) return;
+  clearHistoryEntries({ historyMap: params.historyMap, historyKey: params.historyKey });
+}
+
 export function buildHistoryContextFromEntries(params: {
   entries: HistoryEntry[];
   currentMessage: string;

src/auto-reply/reply/provider-dispatcher.ts

@@ -1,58 +1,44 @@
 import type { ClawdbotConfig } from "../../config/config.js";
-import type { FinalizedMsgContext } from "../templating.js";
+import type { FinalizedMsgContext, MsgContext } from "../templating.js";
 import type { GetReplyOptions } from "../types.js";
-import type { DispatchFromConfigResult } from "./dispatch-from-config.js";
-import { dispatchReplyFromConfig } from "./dispatch-from-config.js";
+import type { DispatchInboundResult } from "../dispatch.js";
 import {
-  createReplyDispatcher,
-  createReplyDispatcherWithTyping,
-  type ReplyDispatcherOptions,
-  type ReplyDispatcherWithTypingOptions,
+  dispatchInboundMessageWithBufferedDispatcher,
+  dispatchInboundMessageWithDispatcher,
+} from "../dispatch.js";
+import type {
+  ReplyDispatcherOptions,
+  ReplyDispatcherWithTypingOptions,
 } from "./reply-dispatcher.js";
 
 export async function dispatchReplyWithBufferedBlockDispatcher(params: {
-  ctx: FinalizedMsgContext;
+  ctx: MsgContext | FinalizedMsgContext;
   cfg: ClawdbotConfig;
   dispatcherOptions: ReplyDispatcherWithTypingOptions;
   replyOptions?: Omit<GetReplyOptions, "onToolResult" | "onBlockReply">;
   replyResolver?: typeof import("../reply.js").getReplyFromConfig;
-}): Promise<DispatchFromConfigResult> {
-  const { dispatcher, replyOptions, markDispatchIdle } = createReplyDispatcherWithTyping(
-    params.dispatcherOptions,
-  );
-
-  const result = await dispatchReplyFromConfig({
+}): Promise<DispatchInboundResult> {
+  return await dispatchInboundMessageWithBufferedDispatcher({
     ctx: params.ctx,
     cfg: params.cfg,
-    dispatcher,
+    dispatcherOptions: params.dispatcherOptions,
     replyResolver: params.replyResolver,
-    replyOptions: {
-      ...params.replyOptions,
-      ...replyOptions,
-    },
+    replyOptions: params.replyOptions,
   });
-
-  markDispatchIdle();
-  return result;
 }
 
 export async function dispatchReplyWithDispatcher(params: {
-  ctx: FinalizedMsgContext;
+  ctx: MsgContext | FinalizedMsgContext;
   cfg: ClawdbotConfig;
   dispatcherOptions: ReplyDispatcherOptions;
   replyOptions?: Omit<GetReplyOptions, "onToolResult" | "onBlockReply">;
   replyResolver?: typeof import("../reply.js").getReplyFromConfig;
-}): Promise<DispatchFromConfigResult> {
-  const dispatcher = createReplyDispatcher(params.dispatcherOptions);
-
-  const result = await dispatchReplyFromConfig({
+}): Promise<DispatchInboundResult> {
+  return await dispatchInboundMessageWithDispatcher({
     ctx: params.ctx,
     cfg: params.cfg,
-    dispatcher,
+    dispatcherOptions: params.dispatcherOptions,
     replyResolver: params.replyResolver,
     replyOptions: params.replyOptions,
   });
-
-  await dispatcher.waitForIdle();
-  return result;
 }

src/auto-reply/types.ts

@@ -1,3 +1,4 @@
+import type { ImageContent } from "@mariozechner/pi-ai";
 import type { TypingController } from "./reply/typing.js";
 
 export type BlockReplyContext = {
@@ -13,6 +14,14 @@ export type ModelSelectedContext = {
 };
 
 export type GetReplyOptions = {
+  /** Override run id for agent events (defaults to random UUID). */
+  runId?: string;
+  /** Abort signal for the underlying agent run. */
+  abortSignal?: AbortSignal;
+  /** Optional inbound images (used for webchat attachments). */
+  images?: ImageContent[];
+  /** Notifies when an agent run actually starts (useful for webchat command handling). */
+  onAgentRunStart?: (runId: string) => void;
   onReplyStart?: () => Promise<void> | void;
   onTypingController?: (typing: TypingController) => void;
   isHeartbeat?: boolean;

src/channels/ack-reactions.test.ts

@@ -0,0 +1,269 @@
+import { describe, expect, it, vi } from "vitest";
+
+import {
+  removeAckReactionAfterReply,
+  shouldAckReaction,
+  shouldAckReactionForWhatsApp,
+} from "./ack-reactions.js";
+
+describe("shouldAckReaction", () => {
+  it("honors direct and group-all scopes", () => {
+    expect(
+      shouldAckReaction({
+        scope: "direct",
+        isDirect: true,
+        isGroup: false,
+        isMentionableGroup: false,
+        requireMention: false,
+        canDetectMention: false,
+        effectiveWasMentioned: false,
+      }),
+    ).toBe(true);
+
+    expect(
+      shouldAckReaction({
+        scope: "group-all",
+        isDirect: false,
+        isGroup: true,
+        isMentionableGroup: true,
+        requireMention: false,
+        canDetectMention: false,
+        effectiveWasMentioned: false,
+      }),
+    ).toBe(true);
+  });
+
+  it("skips when scope is off or none", () => {
+    expect(
+      shouldAckReaction({
+        scope: "off",
+        isDirect: true,
+        isGroup: true,
+        isMentionableGroup: true,
+        requireMention: true,
+        canDetectMention: true,
+        effectiveWasMentioned: true,
+      }),
+    ).toBe(false);
+
+    expect(
+      shouldAckReaction({
+        scope: "none",
+        isDirect: true,
+        isGroup: true,
+        isMentionableGroup: true,
+        requireMention: true,
+        canDetectMention: true,
+        effectiveWasMentioned: true,
+      }),
+    ).toBe(false);
+  });
+
+  it("defaults to group-mentions gating", () => {
+    expect(
+      shouldAckReaction({
+        scope: undefined,
+        isDirect: false,
+        isGroup: true,
+        isMentionableGroup: true,
+        requireMention: true,
+        canDetectMention: true,
+        effectiveWasMentioned: true,
+      }),
+    ).toBe(true);
+  });
+
+  it("requires mention gating for group-mentions", () => {
+    expect(
+      shouldAckReaction({
+        scope: "group-mentions",
+        isDirect: false,
+        isGroup: true,
+        isMentionableGroup: true,
+        requireMention: false,
+        canDetectMention: true,
+        effectiveWasMentioned: true,
+      }),
+    ).toBe(false);
+
+    expect(
+      shouldAckReaction({
+        scope: "group-mentions",
+        isDirect: false,
+        isGroup: true,
+        isMentionableGroup: true,
+        requireMention: true,
+        canDetectMention: false,
+        effectiveWasMentioned: true,
+      }),
+    ).toBe(false);
+
+    expect(
+      shouldAckReaction({
+        scope: "group-mentions",
+        isDirect: false,
+        isGroup: true,
+        isMentionableGroup: false,
+        requireMention: true,
+        canDetectMention: true,
+        effectiveWasMentioned: true,
+      }),
+    ).toBe(false);
+
+    expect(
+      shouldAckReaction({
+        scope: "group-mentions",
+        isDirect: false,
+        isGroup: true,
+        isMentionableGroup: true,
+        requireMention: true,
+        canDetectMention: true,
+        effectiveWasMentioned: true,
+      }),
+    ).toBe(true);
+
+    expect(
+      shouldAckReaction({
+        scope: "group-mentions",
+        isDirect: false,
+        isGroup: true,
+        isMentionableGroup: true,
+        requireMention: true,
+        canDetectMention: true,
+        effectiveWasMentioned: false,
+        shouldBypassMention: true,
+      }),
+    ).toBe(true);
+  });
+});
+
+describe("shouldAckReactionForWhatsApp", () => {
+  it("respects direct and group modes", () => {
+    expect(
+      shouldAckReactionForWhatsApp({
+        emoji: "👀",
+        isDirect: true,
+        isGroup: false,
+        directEnabled: true,
+        groupMode: "mentions",
+        wasMentioned: false,
+        groupActivated: false,
+      }),
+    ).toBe(true);
+
+    expect(
+      shouldAckReactionForWhatsApp({
+        emoji: "👀",
+        isDirect: true,
+        isGroup: false,
+        directEnabled: false,
+        groupMode: "mentions",
+        wasMentioned: false,
+        groupActivated: false,
+      }),
+    ).toBe(false);
+
+    expect(
+      shouldAckReactionForWhatsApp({
+        emoji: "👀",
+        isDirect: false,
+        isGroup: true,
+        directEnabled: true,
+        groupMode: "always",
+        wasMentioned: false,
+        groupActivated: false,
+      }),
+    ).toBe(true);
+
+    expect(
+      shouldAckReactionForWhatsApp({
+        emoji: "👀",
+        isDirect: false,
+        isGroup: true,
+        directEnabled: true,
+        groupMode: "never",
+        wasMentioned: true,
+        groupActivated: true,
+      }),
+    ).toBe(false);
+  });
+
+  it("honors mentions or activation for group-mentions", () => {
+    expect(
+      shouldAckReactionForWhatsApp({
+        emoji: "👀",
+        isDirect: false,
+        isGroup: true,
+        directEnabled: true,
+        groupMode: "mentions",
+        wasMentioned: true,
+        groupActivated: false,
+      }),
+    ).toBe(true);
+
+    expect(
+      shouldAckReactionForWhatsApp({
+        emoji: "👀",
+        isDirect: false,
+        isGroup: true,
+        directEnabled: true,
+        groupMode: "mentions",
+        wasMentioned: false,
+        groupActivated: true,
+      }),
+    ).toBe(true);
+
+    expect(
+      shouldAckReactionForWhatsApp({
+        emoji: "👀",
+        isDirect: false,
+        isGroup: true,
+        directEnabled: true,
+        groupMode: "mentions",
+        wasMentioned: false,
+        groupActivated: false,
+      }),
+    ).toBe(false);
+  });
+});
+
+describe("removeAckReactionAfterReply", () => {
+  it("removes only when ack succeeded", async () => {
+    const remove = vi.fn().mockResolvedValue(undefined);
+    const onError = vi.fn();
+    removeAckReactionAfterReply({
+      removeAfterReply: true,
+      ackReactionPromise: Promise.resolve(true),
+      ackReactionValue: "👀",
+      remove,
+      onError,
+    });
+    await new Promise((resolve) => setTimeout(resolve, 0));
+    expect(remove).toHaveBeenCalledTimes(1);
+    expect(onError).not.toHaveBeenCalled();
+  });
+
+  it("skips removal when ack did not happen", async () => {
+    const remove = vi.fn().mockResolvedValue(undefined);
+    removeAckReactionAfterReply({
+      removeAfterReply: true,
+      ackReactionPromise: Promise.resolve(false),
+      ackReactionValue: "👀",
+      remove,
+    });
+    await new Promise((resolve) => setTimeout(resolve, 0));
+    expect(remove).not.toHaveBeenCalled();
+  });
+
+  it("skips when not configured", async () => {
+    const remove = vi.fn().mockResolvedValue(undefined);
+    removeAckReactionAfterReply({
+      removeAfterReply: false,
+      ackReactionPromise: Promise.resolve(true),
+      ackReactionValue: "👀",
+      remove,
+    });
+    await new Promise((resolve) => setTimeout(resolve, 0));
+    expect(remove).not.toHaveBeenCalled();
+  });
+});

src/channels/ack-reactions.ts

@@ -0,0 +1,71 @@
+export type AckReactionScope = "all" | "direct" | "group-all" | "group-mentions" | "off" | "none";
+
+export type WhatsAppAckReactionMode = "always" | "mentions" | "never";
+
+export type AckReactionGateParams = {
+  scope: AckReactionScope | undefined;
+  isDirect: boolean;
+  isGroup: boolean;
+  isMentionableGroup: boolean;
+  requireMention: boolean;
+  canDetectMention: boolean;
+  effectiveWasMentioned: boolean;
+  shouldBypassMention?: boolean;
+};
+
+export function shouldAckReaction(params: AckReactionGateParams): boolean {
+  const scope = params.scope ?? "group-mentions";
+  if (scope === "off" || scope === "none") return false;
+  if (scope === "all") return true;
+  if (scope === "direct") return params.isDirect;
+  if (scope === "group-all") return params.isGroup;
+  if (scope === "group-mentions") {
+    if (!params.isMentionableGroup) return false;
+    if (!params.requireMention) return false;
+    if (!params.canDetectMention) return false;
+    return params.effectiveWasMentioned || params.shouldBypassMention === true;
+  }
+  return false;
+}
+
+export function shouldAckReactionForWhatsApp(params: {
+  emoji: string;
+  isDirect: boolean;
+  isGroup: boolean;
+  directEnabled: boolean;
+  groupMode: WhatsAppAckReactionMode;
+  wasMentioned: boolean;
+  groupActivated: boolean;
+}): boolean {
+  if (!params.emoji) return false;
+  if (params.isDirect) return params.directEnabled;
+  if (!params.isGroup) return false;
+  if (params.groupMode === "never") return false;
+  if (params.groupMode === "always") return true;
+  return shouldAckReaction({
+    scope: "group-mentions",
+    isDirect: false,
+    isGroup: true,
+    isMentionableGroup: true,
+    requireMention: true,
+    canDetectMention: true,
+    effectiveWasMentioned: params.wasMentioned,
+    shouldBypassMention: params.groupActivated,
+  });
+}
+
+export function removeAckReactionAfterReply(params: {
+  removeAfterReply: boolean;
+  ackReactionPromise: Promise<boolean> | null;
+  ackReactionValue: string | null;
+  remove: () => Promise<void>;
+  onError?: (err: unknown) => void;
+}) {
+  if (!params.removeAfterReply) return;
+  if (!params.ackReactionPromise) return;
+  if (!params.ackReactionValue) return;
+  void params.ackReactionPromise.then((didAck) => {
+    if (!didAck) return;
+    params.remove().catch((err) => params.onError?.(err));
+  });
+}

src/channels/logging.ts

@@ -0,0 +1,33 @@
+export type LogFn = (message: string) => void;
+
+export function logInboundDrop(params: {
+  log: LogFn;
+  channel: string;
+  reason: string;
+  target?: string;
+}): void {
+  const target = params.target ? ` target=${params.target}` : "";
+  params.log(`${params.channel}: drop ${params.reason}${target}`);
+}
+
+export function logTypingFailure(params: {
+  log: LogFn;
+  channel: string;
+  target?: string;
+  action?: "start" | "stop";
+  error: unknown;
+}): void {
+  const target = params.target ? ` target=${params.target}` : "";
+  const action = params.action ? ` action=${params.action}` : "";
+  params.log(`${params.channel} typing${action} failed${target}: ${String(params.error)}`);
+}
+
+export function logAckFailure(params: {
+  log: LogFn;
+  channel: string;
+  target?: string;
+  error: unknown;
+}): void {
+  const target = params.target ? ` target=${params.target}` : "";
+  params.log(`${params.channel} ack cleanup failed${target}: ${String(params.error)}`);
+}

src/channels/reply-prefix.ts

@@ -0,0 +1,41 @@
+import { resolveEffectiveMessagesConfig, resolveIdentityName } from "../agents/identity.js";
+import type { ClawdbotConfig } from "../config/config.js";
+import type { GetReplyOptions } from "../auto-reply/types.js";
+import {
+  extractShortModelName,
+  type ResponsePrefixContext,
+} from "../auto-reply/reply/response-prefix-template.js";
+
+type ModelSelectionContext = Parameters<NonNullable<GetReplyOptions["onModelSelected"]>>[0];
+
+export type ReplyPrefixContextBundle = {
+  prefixContext: ResponsePrefixContext;
+  responsePrefix?: string;
+  responsePrefixContextProvider: () => ResponsePrefixContext;
+  onModelSelected: (ctx: ModelSelectionContext) => void;
+};
+
+export function createReplyPrefixContext(params: {
+  cfg: ClawdbotConfig;
+  agentId: string;
+}): ReplyPrefixContextBundle {
+  const { cfg, agentId } = params;
+  const prefixContext: ResponsePrefixContext = {
+    identityName: resolveIdentityName(cfg, agentId),
+  };
+
+  const onModelSelected = (ctx: ModelSelectionContext) => {
+    // Mutate the object directly instead of reassigning to ensure closures see updates.
+    prefixContext.provider = ctx.provider;
+    prefixContext.model = extractShortModelName(ctx.model);
+    prefixContext.modelFull = `${ctx.provider}/${ctx.model}`;
+    prefixContext.thinkingLevel = ctx.thinkLevel ?? "off";
+  };
+
+  return {
+    prefixContext,
+    responsePrefix: resolveEffectiveMessagesConfig(cfg, agentId).responsePrefix,
+    responsePrefixContextProvider: () => prefixContext,
+    onModelSelected,
+  };
+}

src/channels/session.ts

@@ -0,0 +1,49 @@
+import type { MsgContext } from "../auto-reply/templating.js";
+import {
+  recordSessionMetaFromInbound,
+  type GroupKeyResolution,
+  type SessionEntry,
+  updateLastRoute,
+} from "../config/sessions.js";
+
+export type InboundLastRouteUpdate = {
+  sessionKey: string;
+  channel: SessionEntry["lastChannel"];
+  to: string;
+  accountId?: string;
+  threadId?: string | number;
+};
+
+export async function recordInboundSession(params: {
+  storePath: string;
+  sessionKey: string;
+  ctx: MsgContext;
+  groupResolution?: GroupKeyResolution | null;
+  createIfMissing?: boolean;
+  updateLastRoute?: InboundLastRouteUpdate;
+  onRecordError: (err: unknown) => void;
+}): Promise<void> {
+  const { storePath, sessionKey, ctx, groupResolution, createIfMissing } = params;
+  void recordSessionMetaFromInbound({
+    storePath,
+    sessionKey,
+    ctx,
+    groupResolution,
+    createIfMissing,
+  }).catch(params.onRecordError);
+
+  const update = params.updateLastRoute;
+  if (!update) return;
+  await updateLastRoute({
+    storePath,
+    sessionKey: update.sessionKey,
+    deliveryContext: {
+      channel: update.channel,
+      to: update.to,
+      accountId: update.accountId,
+      threadId: update.threadId,
+    },
+    ctx,
+    groupResolution,
+  });
+}

src/channels/typing.test.ts

@@ -0,0 +1,42 @@
+import { describe, expect, it, vi } from "vitest";
+
+import { createTypingCallbacks } from "./typing.js";
+
+const flush = () => new Promise((resolve) => setTimeout(resolve, 0));
+
+describe("createTypingCallbacks", () => {
+  it("invokes start on reply start", async () => {
+    const start = vi.fn().mockResolvedValue(undefined);
+    const onStartError = vi.fn();
+    const callbacks = createTypingCallbacks({ start, onStartError });
+
+    await callbacks.onReplyStart();
+
+    expect(start).toHaveBeenCalledTimes(1);
+    expect(onStartError).not.toHaveBeenCalled();
+  });
+
+  it("reports start errors", async () => {
+    const start = vi.fn().mockRejectedValue(new Error("fail"));
+    const onStartError = vi.fn();
+    const callbacks = createTypingCallbacks({ start, onStartError });
+
+    await callbacks.onReplyStart();
+
+    expect(onStartError).toHaveBeenCalledTimes(1);
+  });
+
+  it("invokes stop on idle and reports stop errors", async () => {
+    const start = vi.fn().mockResolvedValue(undefined);
+    const stop = vi.fn().mockRejectedValue(new Error("stop"));
+    const onStartError = vi.fn();
+    const onStopError = vi.fn();
+    const callbacks = createTypingCallbacks({ start, stop, onStartError, onStopError });
+
+    callbacks.onIdle?.();
+    await flush();
+
+    expect(stop).toHaveBeenCalledTimes(1);
+    expect(onStopError).toHaveBeenCalledTimes(1);
+  });
+});

src/channels/typing.ts

@@ -0,0 +1,28 @@
+export type TypingCallbacks = {
+  onReplyStart: () => Promise<void>;
+  onIdle?: () => void;
+};
+
+export function createTypingCallbacks(params: {
+  start: () => Promise<void>;
+  stop?: () => Promise<void>;
+  onStartError: (err: unknown) => void;
+  onStopError?: (err: unknown) => void;
+}): TypingCallbacks {
+  const stop = params.stop;
+  const onReplyStart = async () => {
+    try {
+      await params.start();
+    } catch (err) {
+      params.onStartError(err);
+    }
+  };
+
+  const onIdle = stop
+    ? () => {
+        void stop().catch((err) => (params.onStopError ?? params.onStartError)(err));
+      }
+    : undefined;
+
+  return { onReplyStart, onIdle };
+}

src/commands/doctor-sandbox-image-inspect-errors.test.ts

@@ -0,0 +1,61 @@
+import { beforeEach, describe, expect, it, vi } from "vitest";
+
+import type { ClawdbotConfig } from "../config/config.js";
+import type { RuntimeEnv } from "../runtime.js";
+import type { DoctorPrompter } from "./doctor-prompter.js";
+
+const runExec = vi.fn();
+const runCommandWithTimeout = vi.fn();
+const note = vi.fn();
+
+vi.mock("../process/exec.js", () => ({
+  runExec,
+  runCommandWithTimeout,
+}));
+vi.mock("../terminal/note.js", () => ({
+  note,
+}));
+
+describe("maybeRepairSandboxImages", () => {
+  beforeEach(() => {
+    runExec.mockReset();
+    runCommandWithTimeout.mockReset();
+    note.mockReset();
+  });
+
+  it("logs docker inspect errors and continues", async () => {
+    runExec.mockImplementation(async (_command: string, args: string[]) => {
+      if (args[0] === "version") return { stdout: "26.0.0", stderr: "" };
+      if (args[0] === "image" && args[1] === "inspect") {
+        const err = new Error(
+          "permission denied while trying to connect to the Docker daemon socket",
+        ) as Error & { stderr?: string };
+        err.stderr = "permission denied while trying to connect to the Docker daemon socket";
+        throw err;
+      }
+      return { stdout: "", stderr: "" };
+    });
+
+    const { maybeRepairSandboxImages } = await import("./doctor-sandbox.js");
+    const cfg = {
+      agents: {
+        defaults: {
+          sandbox: {
+            mode: "all",
+            docker: { image: "custom-sandbox:latest" },
+          },
+        },
+      },
+    } satisfies ClawdbotConfig;
+
+    const runtime = { log: vi.fn(), error: vi.fn() } as RuntimeEnv;
+    const prompter = { confirmSkipInNonInteractive: vi.fn() } as DoctorPrompter;
+
+    await expect(maybeRepairSandboxImages(cfg, runtime, prompter)).resolves.toBe(cfg);
+    expect(note).toHaveBeenCalledWith(
+      expect.stringContaining("Unable to inspect sandbox base image (custom-sandbox:latest)"),
+      "Sandbox",
+    );
+    expect(runCommandWithTimeout).not.toHaveBeenCalled();
+  });
+});

src/commands/doctor-sandbox.ts

@@ -78,8 +78,15 @@ async function dockerImageExists(image: string): Promise<boolean> {
   try {
     await runExec("docker", ["image", "inspect", image], { timeoutMs: 5_000 });
     return true;
-  } catch {
-    return false;
+  } catch (error: any) {
+    const stderr = typeof error?.stderr === "string" ? error.stderr.trim() : "";
+    const stdout = typeof error?.stdout === "string" ? error.stdout.trim() : "";
+    const message = typeof error?.message === "string" ? error.message.trim() : "";
+    const detail = stderr || stdout || message;
+    if (/no such image/i.test(detail)) {
+      return false;
+    }
+    throw new Error(`Failed to inspect sandbox image: ${detail || "unknown error"}`);
   }
 }
 
@@ -143,7 +150,17 @@ async function handleMissingSandboxImage(
   runtime: RuntimeEnv,
   prompter: DoctorPrompter,
 ) {
-  const exists = await dockerImageExists(params.image);
+  let exists = false;
+  try {
+    exists = await dockerImageExists(params.image);
+  } catch (error: any) {
+    const stderr = typeof error?.stderr === "string" ? error.stderr.trim() : "";
+    const stdout = typeof error?.stdout === "string" ? error.stdout.trim() : "";
+    const message = typeof error?.message === "string" ? error.message.trim() : "";
+    const detail = stderr || stdout || message || "unknown error";
+    note(`Unable to inspect sandbox ${params.label} image (${params.image}): ${detail}`, "Sandbox");
+    return;
+  }
   if (exists) return;
 
   const buildHint = params.buildScript

src/commands/models/list.status-command.ts

@@ -28,6 +28,7 @@ import {
 } from "../../infra/provider-usage.js";
 import type { RuntimeEnv } from "../../runtime.js";
 import { colorize, theme } from "../../terminal/theme.js";
+import { renderTable } from "../../terminal/table.js";
 import { formatCliCommand } from "../../cli/command-format.js";
 import { shortenHomePath } from "../../utils.js";
 import { resolveProviderAuthOverview } from "./list.auth-overview.js";
@@ -35,7 +36,6 @@ import { isRich } from "./list.format.js";
 import {
   describeProbeSummary,
   formatProbeLatency,
-  groupProbeResults,
   runAuthProbes,
   sortProbeResults,
   type AuthProbeSummary,
@@ -571,7 +571,8 @@ export async function modelsStatusCommand(
     if (probeSummary.results.length === 0) {
       runtime.log(colorize(rich, theme.muted, "- none"));
     } else {
-      const grouped = groupProbeResults(sortProbeResults(probeSummary.results));
+      const tableWidth = Math.max(60, (process.stdout.columns ?? 120) - 1);
+      const sorted = sortProbeResults(probeSummary.results);
       const statusColor = (status: string) => {
         if (status === "ok") return theme.success;
         if (status === "rate_limit") return theme.warn;
@@ -580,29 +581,33 @@ export async function modelsStatusCommand(
         if (status === "no_model") return theme.muted;
         return theme.muted;
       };
-      for (const [provider, results] of grouped) {
-        const modelLabel = results.find((r) => r.model)?.model ?? "-";
-        runtime.log(
-          `- ${theme.heading(provider)}${colorize(
-            rich,
-            theme.muted,
-            modelLabel ? ` (model: ${modelLabel})` : "",
-          )}`,
-        );
-        for (const result of results) {
-          const status = colorize(rich, statusColor(result.status), result.status);
-          const latency = formatProbeLatency(result.latencyMs);
-          const mode = result.mode ? ` (${result.mode})` : "";
-          const detail = result.error ? colorize(rich, theme.muted, ` - ${result.error}`) : "";
-          runtime.log(
-            `  - ${colorize(rich, theme.accent, result.label)}${mode} ${status} ${colorize(
-              rich,
-              theme.muted,
-              latency,
-            )}${detail}`,
-          );
-        }
-      }
+      const rows = sorted.map((result) => {
+        const status = colorize(rich, statusColor(result.status), result.status);
+        const latency = formatProbeLatency(result.latencyMs);
+        const detail = result.error ? colorize(rich, theme.muted, result.error) : "";
+        const modelLabel = result.model ?? `${result.provider}/-`;
+        const modeLabel = result.mode ? ` ${colorize(rich, theme.muted, `(${result.mode})`)}` : "";
+        const profile = `${colorize(rich, theme.accent, result.label)}${modeLabel}`;
+        const statusLabel = `${status}${colorize(rich, theme.muted, ` · ${latency}`)}`;
+        return {
+          Model: colorize(rich, theme.heading, modelLabel),
+          Profile: profile,
+          Status: statusLabel,
+          Detail: detail,
+        };
+      });
+      runtime.log(
+        renderTable({
+          width: tableWidth,
+          columns: [
+            { key: "Model", header: "Model", minWidth: 18 },
+            { key: "Profile", header: "Profile", minWidth: 24 },
+            { key: "Status", header: "Status", minWidth: 12 },
+            { key: "Detail", header: "Detail", minWidth: 16, flex: true },
+          ],
+          rows,
+        }).trimEnd(),
+      );
       runtime.log(colorize(rich, theme.muted, describeProbeSummary(probeSummary)));
     }
   }

src/discord/monitor.slash.test.ts

@@ -1,4 +1,5 @@
 import { beforeEach, describe, expect, it, vi } from "vitest";
+import { createReplyDispatcherWithTyping } from "../auto-reply/reply/reply-dispatcher.js";
 
 const dispatchMock = vi.fn();
 
@@ -20,15 +21,34 @@ vi.mock("@buape/carbon", () => ({
   },
 }));
 
-vi.mock("../auto-reply/reply/dispatch-from-config.js", () => ({
-  dispatchReplyFromConfig: (...args: unknown[]) => dispatchMock(...args),
-}));
+vi.mock("../auto-reply/dispatch.js", async (importOriginal) => {
+  const actual = await importOriginal<typeof import("../auto-reply/dispatch.js")>();
+  return {
+    ...actual,
+    dispatchInboundMessage: (...args: unknown[]) => dispatchMock(...args),
+    dispatchInboundMessageWithDispatcher: (...args: unknown[]) => dispatchMock(...args),
+    dispatchInboundMessageWithBufferedDispatcher: (...args: unknown[]) => dispatchMock(...args),
+  };
+});
 
 beforeEach(() => {
-  dispatchMock.mockReset().mockImplementation(async ({ dispatcher }) => {
-    dispatcher.sendToolResult({ text: "tool update" });
-    dispatcher.sendFinalReply({ text: "final reply" });
-    return { queuedFinal: true, counts: { tool: 1, block: 0, final: 1 } };
+  dispatchMock.mockReset().mockImplementation(async (params) => {
+    if ("dispatcher" in params && params.dispatcher) {
+      params.dispatcher.sendToolResult({ text: "tool update" });
+      params.dispatcher.sendFinalReply({ text: "final reply" });
+      return { queuedFinal: true, counts: { tool: 1, block: 0, final: 1 } };
+    }
+    if ("dispatcherOptions" in params && params.dispatcherOptions) {
+      const { dispatcher, markDispatchIdle } = createReplyDispatcherWithTyping(
+        params.dispatcherOptions,
+      );
+      dispatcher.sendToolResult({ text: "tool update" });
+      dispatcher.sendFinalReply({ text: "final reply" });
+      await dispatcher.waitForIdle();
+      markDispatchIdle();
+      return { queuedFinal: true, counts: dispatcher.getQueuedCounts() };
+    }
+    return { queuedFinal: false, counts: { tool: 0, block: 0, final: 0 } };
   });
 });
 

src/discord/monitor.tool-result.accepts-guild-messages-mentionpatterns-match.test.ts

@@ -18,9 +18,15 @@ vi.mock("./send.js", () => ({
     reactMock(...args);
   },
 }));
-vi.mock("../auto-reply/reply/dispatch-from-config.js", () => ({
-  dispatchReplyFromConfig: (...args: unknown[]) => dispatchMock(...args),
-}));
+vi.mock("../auto-reply/dispatch.js", async (importOriginal) => {
+  const actual = await importOriginal<typeof import("../auto-reply/dispatch.js")>();
+  return {
+    ...actual,
+    dispatchInboundMessage: (...args: unknown[]) => dispatchMock(...args),
+    dispatchInboundMessageWithDispatcher: (...args: unknown[]) => dispatchMock(...args),
+    dispatchInboundMessageWithBufferedDispatcher: (...args: unknown[]) => dispatchMock(...args),
+  };
+});
 vi.mock("../pairing/pairing-store.js", () => ({
   readChannelAllowFromStore: (...args: unknown[]) => readAllowFromStoreMock(...args),
   upsertChannelPairingRequest: (...args: unknown[]) => upsertPairingRequestMock(...args),
@@ -41,7 +47,7 @@ beforeEach(() => {
   updateLastRouteMock.mockReset();
   dispatchMock.mockReset().mockImplementation(async ({ dispatcher }) => {
     dispatcher.sendFinalReply({ text: "hi" });
-    return { queuedFinal: true, counts: { final: 1 } };
+    return { queuedFinal: true, counts: { tool: 0, block: 0, final: 1 } };
   });
   readAllowFromStoreMock.mockReset().mockResolvedValue([]);
   upsertPairingRequestMock.mockReset().mockResolvedValue({ code: "PAIRCODE", created: true });

src/discord/monitor.tool-result.sends-status-replies-responseprefix.test.ts

@@ -18,9 +18,15 @@ vi.mock("./send.js", () => ({
     reactMock(...args);
   },
 }));
-vi.mock("../auto-reply/reply/dispatch-from-config.js", () => ({
-  dispatchReplyFromConfig: (...args: unknown[]) => dispatchMock(...args),
-}));
+vi.mock("../auto-reply/dispatch.js", async (importOriginal) => {
+  const actual = await importOriginal<typeof import("../auto-reply/dispatch.js")>();
+  return {
+    ...actual,
+    dispatchInboundMessage: (...args: unknown[]) => dispatchMock(...args),
+    dispatchInboundMessageWithDispatcher: (...args: unknown[]) => dispatchMock(...args),
+    dispatchInboundMessageWithBufferedDispatcher: (...args: unknown[]) => dispatchMock(...args),
+  };
+});
 vi.mock("../pairing/pairing-store.js", () => ({
   readChannelAllowFromStore: (...args: unknown[]) => readAllowFromStoreMock(...args),
   upsertChannelPairingRequest: (...args: unknown[]) => upsertPairingRequestMock(...args),
@@ -40,7 +46,7 @@ beforeEach(() => {
   updateLastRouteMock.mockReset();
   dispatchMock.mockReset().mockImplementation(async ({ dispatcher }) => {
     dispatcher.sendFinalReply({ text: "hi" });
-    return { queuedFinal: true, counts: { final: 1 } };
+    return { queuedFinal: true, counts: { tool: 0, block: 0, final: 1 } };
   });
   readAllowFromStoreMock.mockReset().mockResolvedValue([]);
   upsertPairingRequestMock.mockReset().mockResolvedValue({ code: "PAIRCODE", created: true });

src/discord/monitor/message-handler.inbound-contract.test.ts

@@ -9,17 +9,24 @@ import { expectInboundContextContract } from "../../../test/helpers/inbound-cont
 
 let capturedCtx: MsgContext | undefined;
 
-vi.mock("../../auto-reply/reply/dispatch-from-config.js", () => ({
-  dispatchReplyFromConfig: vi.fn(async (params: { ctx: MsgContext }) => {
+vi.mock("../../auto-reply/dispatch.js", async (importOriginal) => {
+  const actual = await importOriginal<typeof import("../../auto-reply/dispatch.js")>();
+  const dispatchInboundMessage = vi.fn(async (params: { ctx: MsgContext }) => {
     capturedCtx = params.ctx;
-    return { queuedFinal: false, counts: { tool: 0, block: 0 } };
-  }),
-}));
+    return { queuedFinal: false, counts: { tool: 0, block: 0, final: 0 } };
+  });
+  return {
+    ...actual,
+    dispatchInboundMessage,
+    dispatchInboundMessageWithDispatcher: dispatchInboundMessage,
+    dispatchInboundMessageWithBufferedDispatcher: dispatchInboundMessage,
+  };
+});
 
 import { processDiscordMessage } from "./message-handler.process.js";
 
 describe("discord processDiscordMessage inbound contract", () => {
-  it("passes a finalized MsgContext to dispatchReplyFromConfig", async () => {
+  it("passes a finalized MsgContext to dispatchInboundMessage", async () => {
     capturedCtx = undefined;
 
     const dir = await fs.mkdtemp(path.join(os.tmpdir(), "clawdbot-discord-"));

src/discord/monitor/message-handler.preflight.ts

@@ -2,7 +2,10 @@ import { ChannelType, MessageType, type User } from "@buape/carbon";
 
 import { hasControlCommand } from "../../auto-reply/command-detection.js";
 import { shouldHandleTextCommands } from "../../auto-reply/commands-registry.js";
-import { recordPendingHistoryEntry, type HistoryEntry } from "../../auto-reply/reply/history.js";
+import {
+  recordPendingHistoryEntryIfEnabled,
+  type HistoryEntry,
+} from "../../auto-reply/reply/history.js";
 import { buildMentionRegexes, matchesMentionPatterns } from "../../auto-reply/reply/mentions.js";
 import { logVerbose, shouldLogVerbose } from "../../globals.js";
 import { recordChannelActivity } from "../../infra/channel-activity.js";
@@ -18,6 +21,7 @@ import { resolveMentionGatingWithBypass } from "../../channels/mention-gating.js
 import { formatAllowlistMatchMeta } from "../../channels/allowlist-match.js";
 import { sendMessageDiscord } from "../send.js";
 import { resolveControlCommandGate } from "../../channels/command-gating.js";
+import { logInboundDrop } from "../../channels/logging.js";
 import {
   allowListMatches,
   isDiscordGroupAllowedByPolicy,
@@ -382,7 +386,12 @@ export async function preflightDiscordMessage(
     commandAuthorized = commandGate.commandAuthorized;
 
     if (commandGate.shouldBlock) {
-      logVerbose(`Blocked discord control command from unauthorized sender ${author.id}`);
+      logInboundDrop({
+        log: logVerbose,
+        channel: "discord",
+        reason: "control command (unauthorized)",
+        target: author.id,
+      });
       return null;
     }
   }
@@ -410,14 +419,12 @@ export async function preflightDiscordMessage(
         },
         "discord: skipping guild message",
       );
-      if (historyEntry && params.historyLimit > 0) {
-        recordPendingHistoryEntry({
-          historyMap: params.guildHistories,
-          historyKey: message.channelId,
-          limit: params.historyLimit,
-          entry: historyEntry,
-        });
-      }
+      recordPendingHistoryEntryIfEnabled({
+        historyMap: params.guildHistories,
+        historyKey: message.channelId,
+        limit: params.historyLimit,
+        entry: historyEntry ?? null,
+      });
       return null;
     }
   }

src/discord/monitor/message-handler.process.ts

@@ -1,32 +1,26 @@
+import { resolveAckReaction, resolveHumanDelayConfig } from "../../agents/identity.js";
 import {
-  resolveAckReaction,
-  resolveEffectiveMessagesConfig,
-  resolveHumanDelayConfig,
-  resolveIdentityName,
-} from "../../agents/identity.js";
-import {
-  extractShortModelName,
-  type ResponsePrefixContext,
-} from "../../auto-reply/reply/response-prefix-template.js";
+  removeAckReactionAfterReply,
+  shouldAckReaction as shouldAckReactionGate,
+} from "../../channels/ack-reactions.js";
+import { logTypingFailure, logAckFailure } from "../../channels/logging.js";
+import { createReplyPrefixContext } from "../../channels/reply-prefix.js";
+import { createTypingCallbacks } from "../../channels/typing.js";
 import {
   formatInboundEnvelope,
   formatThreadStarterEnvelope,
   resolveEnvelopeFormatOptions,
 } from "../../auto-reply/envelope.js";
-import { dispatchReplyFromConfig } from "../../auto-reply/reply/dispatch-from-config.js";
+import { dispatchInboundMessage } from "../../auto-reply/dispatch.js";
 import {
   buildPendingHistoryContextFromMap,
-  clearHistoryEntries,
+  clearHistoryEntriesIfEnabled,
 } from "../../auto-reply/reply/history.js";
 import { finalizeInboundContext } from "../../auto-reply/reply/inbound-context.js";
 import { createReplyDispatcherWithTyping } from "../../auto-reply/reply/reply-dispatcher.js";
 import type { ReplyPayload } from "../../auto-reply/types.js";
-import {
-  readSessionUpdatedAt,
-  recordSessionMetaFromInbound,
-  resolveStorePath,
-  updateLastRoute,
-} from "../../config/sessions.js";
+import { recordInboundSession } from "../../channels/session.js";
+import { readSessionUpdatedAt, resolveStorePath } from "../../config/sessions.js";
 import { resolveMarkdownTableMode } from "../../config/markdown-tables.js";
 import { danger, logVerbose, shouldLogVerbose } from "../../globals.js";
 import { buildAgentSessionKey } from "../../routing/resolve-route.js";
@@ -73,6 +67,7 @@ export async function processDiscordMessage(ctx: DiscordMessagePreflightContext)
     shouldRequireMention,
     canDetectMention,
     effectiveWasMentioned,
+    shouldBypassMention,
     threadChannel,
     threadParentId,
     threadParentName,
@@ -95,20 +90,20 @@ export async function processDiscordMessage(ctx: DiscordMessagePreflightContext)
   }
   const ackReaction = resolveAckReaction(cfg, route.agentId);
   const removeAckAfterReply = cfg.messages?.removeAckAfterReply ?? false;
-  const shouldAckReaction = () => {
-    if (!ackReaction) return false;
-    if (ackReactionScope === "all") return true;
-    if (ackReactionScope === "direct") return isDirectMessage;
-    const isGroupChat = isGuildMessage || isGroupDm;
-    if (ackReactionScope === "group-all") return isGroupChat;
-    if (ackReactionScope === "group-mentions") {
-      if (!isGuildMessage) return false;
-      if (!shouldRequireMention) return false;
-      if (!canDetectMention) return false;
-      return effectiveWasMentioned;
-    }
-    return false;
-  };
+  const shouldAckReaction = () =>
+    Boolean(
+      ackReaction &&
+      shouldAckReactionGate({
+        scope: ackReactionScope,
+        isDirect: isDirectMessage,
+        isGroup: isGuildMessage || isGroupDm,
+        isMentionableGroup: isGuildMessage,
+        requireMention: Boolean(shouldRequireMention),
+        canDetectMention,
+        effectiveWasMentioned,
+        shouldBypassMention,
+      }),
+    );
   const ackReactionPromise = shouldAckReaction()
     ? reactMessageDiscord(message.channelId, message.id, ackReaction, {
         rest: client.rest,
@@ -288,27 +283,23 @@ export async function processDiscordMessage(ctx: DiscordMessagePreflightContext)
     OriginatingTo: autoThreadContext?.OriginatingTo ?? replyTarget,
   });
 
-  void recordSessionMetaFromInbound({
+  await recordInboundSession({
     storePath,
     sessionKey: ctxPayload.SessionKey ?? route.sessionKey,
     ctx: ctxPayload,
-  }).catch((err) => {
-    logVerbose(`discord: failed updating session meta: ${String(err)}`);
+    updateLastRoute: isDirectMessage
+      ? {
+          sessionKey: route.mainSessionKey,
+          channel: "discord",
+          to: `user:${author.id}`,
+          accountId: route.accountId,
+        }
+      : undefined,
+    onRecordError: (err) => {
+      logVerbose(`discord: failed updating session meta: ${String(err)}`);
+    },
   });
 
-  if (isDirectMessage) {
-    await updateLastRoute({
-      storePath,
-      sessionKey: route.mainSessionKey,
-      deliveryContext: {
-        channel: "discord",
-        to: `user:${author.id}`,
-        accountId: route.accountId,
-      },
-      ctx: ctxPayload,
-    });
-  }
-
   if (shouldLogVerbose()) {
     const preview = truncateUtf16Safe(combinedBody, 200).replace(/\n/g, "\\n");
     logVerbose(
@@ -320,10 +311,7 @@ export async function processDiscordMessage(ctx: DiscordMessagePreflightContext)
     ? deliverTarget.slice("channel:".length)
     : message.channelId;
 
-  // Create mutable context for response prefix template interpolation
-  let prefixContext: ResponsePrefixContext = {
-    identityName: resolveIdentityName(cfg, route.agentId),
-  };
+  const prefixContext = createReplyPrefixContext({ cfg, agentId: route.agentId });
   const tableMode = resolveMarkdownTableMode({
     cfg,
     channel: "discord",
@@ -331,8 +319,8 @@ export async function processDiscordMessage(ctx: DiscordMessagePreflightContext)
   });
 
   const { dispatcher, replyOptions, markDispatchIdle } = createReplyDispatcherWithTyping({
-    responsePrefix: resolveEffectiveMessagesConfig(cfg, route.agentId).responsePrefix,
-    responsePrefixContextProvider: () => prefixContext,
+    responsePrefix: prefixContext.responsePrefix,
+    responsePrefixContextProvider: prefixContext.responsePrefixContextProvider,
     humanDelay: resolveHumanDelayConfig(cfg, route.agentId),
     deliver: async (payload: ReplyPayload) => {
       const replyToId = replyReference.use();
@@ -353,10 +341,20 @@ export async function processDiscordMessage(ctx: DiscordMessagePreflightContext)
     onError: (err, info) => {
       runtime.error?.(danger(`discord ${info.kind} reply failed: ${String(err)}`));
     },
-    onReplyStart: () => sendTyping({ client, channelId: typingChannelId }),
+    onReplyStart: createTypingCallbacks({
+      start: () => sendTyping({ client, channelId: typingChannelId }),
+      onStartError: (err) => {
+        logTypingFailure({
+          log: logVerbose,
+          channel: "discord",
+          target: typingChannelId,
+          error: err,
+        });
+      },
+    }).onReplyStart,
   });
 
-  const { queuedFinal, counts } = await dispatchReplyFromConfig({
+  const { queuedFinal, counts } = await dispatchInboundMessage({
     ctx: ctxPayload,
     cfg,
     dispatcher,
@@ -368,20 +366,17 @@ export async function processDiscordMessage(ctx: DiscordMessagePreflightContext)
           ? !discordConfig.blockStreaming
           : undefined,
       onModelSelected: (ctx) => {
-        // Mutate the object directly instead of reassigning to ensure the closure sees updates
-        prefixContext.provider = ctx.provider;
-        prefixContext.model = extractShortModelName(ctx.model);
-        prefixContext.modelFull = `${ctx.provider}/${ctx.model}`;
-        prefixContext.thinkingLevel = ctx.thinkLevel ?? "off";
+        prefixContext.onModelSelected(ctx);
       },
     },
   });
   markDispatchIdle();
   if (!queuedFinal) {
-    if (isGuildMessage && historyLimit > 0) {
-      clearHistoryEntries({
+    if (isGuildMessage) {
+      clearHistoryEntriesIfEnabled({
         historyMap: guildHistories,
         historyKey: message.channelId,
+        limit: historyLimit,
       });
     }
     return;
@@ -392,23 +387,29 @@ export async function processDiscordMessage(ctx: DiscordMessagePreflightContext)
       `discord: delivered ${finalCount} reply${finalCount === 1 ? "" : "ies"} to ${replyTarget}`,
     );
   }
-  if (removeAckAfterReply && ackReactionPromise && ackReaction) {
-    const ackReactionValue = ackReaction;
-    void ackReactionPromise.then((didAck) => {
-      if (!didAck) return;
-      removeReactionDiscord(message.channelId, message.id, ackReactionValue, {
+  removeAckReactionAfterReply({
+    removeAfterReply: removeAckAfterReply,
+    ackReactionPromise,
+    ackReactionValue: ackReaction,
+    remove: async () => {
+      await removeReactionDiscord(message.channelId, message.id, ackReaction, {
         rest: client.rest,
-      }).catch((err) => {
-        logVerbose(
-          `discord: failed to remove ack reaction from ${message.channelId}/${message.id}: ${String(err)}`,
-        );
       });
-    });
-  }
-  if (isGuildMessage && historyLimit > 0) {
-    clearHistoryEntries({
+    },
+    onError: (err) => {
+      logAckFailure({
+        log: logVerbose,
+        channel: "discord",
+        target: `${message.channelId}/${message.id}`,
+        error: err,
+      });
+    },
+  });
+  if (isGuildMessage) {
+    clearHistoryEntriesIfEnabled({
       historyMap: guildHistories,
       historyKey: message.channelId,
+      limit: historyLimit,
     });
   }
 }

src/discord/monitor/typing.ts

@@ -1,15 +1,9 @@
 import type { Client } from "@buape/carbon";
 
-import { logVerbose } from "../../globals.js";
-
 export async function sendTyping(params: { client: Client; channelId: string }) {
-  try {
-    const channel = await params.client.fetchChannel(params.channelId);
-    if (!channel) return;
-    if ("triggerTyping" in channel && typeof channel.triggerTyping === "function") {
-      await channel.triggerTyping();
-    }
-  } catch (err) {
-    logVerbose(`discord typing cue failed for channel ${params.channelId}: ${String(err)}`);
+  const channel = await params.client.fetchChannel(params.channelId);
+  if (!channel) return;
+  if ("triggerTyping" in channel && typeof channel.triggerTyping === "function") {
+    await channel.triggerTyping();
   }
 }

src/gateway/server-methods/chat.ts

@@ -2,30 +2,18 @@ import { randomUUID } from "node:crypto";
 import fs from "node:fs";
 import path from "node:path";
 
-import { resolveSessionAgentId, resolveAgentWorkspaceDir } from "../../agents/agent-scope.js";
+import { CURRENT_SESSION_VERSION } from "@mariozechner/pi-coding-agent";
+import { resolveSessionAgentId } from "../../agents/agent-scope.js";
+import { resolveEffectiveMessagesConfig, resolveIdentityName } from "../../agents/identity.js";
 import { resolveThinkingDefault } from "../../agents/model-selection.js";
 import { resolveAgentTimeoutMs } from "../../agents/timeout.js";
-import { ensureAgentWorkspace } from "../../agents/workspace.js";
-import { isControlCommandMessage } from "../../auto-reply/command-detection.js";
-import { normalizeCommandBody } from "../../auto-reply/commands-registry.js";
-import { formatInboundEnvelope, resolveEnvelopeFormatOptions } from "../../auto-reply/envelope.js";
-import { buildCommandContext, handleCommands } from "../../auto-reply/reply/commands.js";
-import { parseInlineDirectives } from "../../auto-reply/reply/directive-handling.js";
-import { defaultGroupActivation } from "../../auto-reply/reply/groups.js";
-import { resolveContextTokens } from "../../auto-reply/reply/model-selection.js";
-import { resolveElevatedPermissions } from "../../auto-reply/reply/reply-elevated.js";
+import { dispatchInboundMessage } from "../../auto-reply/dispatch.js";
+import { createReplyDispatcher } from "../../auto-reply/reply/reply-dispatcher.js";
 import {
-  normalizeElevatedLevel,
-  normalizeReasoningLevel,
-  normalizeThinkLevel,
-  normalizeVerboseLevel,
-} from "../../auto-reply/thinking.js";
+  extractShortModelName,
+  type ResponsePrefixContext,
+} from "../../auto-reply/reply/response-prefix-template.js";
 import type { MsgContext } from "../../auto-reply/templating.js";
-import { agentCommand } from "../../commands/agent.js";
-import { mergeSessionEntry, updateSessionStore } from "../../config/sessions.js";
-import { registerAgentRunContext } from "../../infra/agent-events.js";
-import { isAcpSessionKey } from "../../routing/session-key.js";
-import { defaultRuntime } from "../../runtime.js";
 import { resolveSendPolicy } from "../../sessions/send-policy.js";
 import { INTERNAL_MESSAGE_CHANNEL } from "../../utils/message-channel.js";
 import {
@@ -53,7 +41,144 @@ import {
 } from "../session-utils.js";
 import { stripEnvelopeFromMessages } from "../chat-sanitize.js";
 import { formatForLog } from "../ws-log.js";
-import type { GatewayRequestHandlers } from "./types.js";
+import type { GatewayRequestContext, GatewayRequestHandlers } from "./types.js";
+
+type TranscriptAppendResult = {
+  ok: boolean;
+  messageId?: string;
+  message?: Record<string, unknown>;
+  error?: string;
+};
+
+function resolveTranscriptPath(params: {
+  sessionId: string;
+  storePath: string | undefined;
+  sessionFile?: string;
+}): string | null {
+  const { sessionId, storePath, sessionFile } = params;
+  if (sessionFile) return sessionFile;
+  if (!storePath) return null;
+  return path.join(path.dirname(storePath), `${sessionId}.jsonl`);
+}
+
+function ensureTranscriptFile(params: { transcriptPath: string; sessionId: string }): {
+  ok: boolean;
+  error?: string;
+} {
+  if (fs.existsSync(params.transcriptPath)) return { ok: true };
+  try {
+    fs.mkdirSync(path.dirname(params.transcriptPath), { recursive: true });
+    const header = {
+      type: "session",
+      version: CURRENT_SESSION_VERSION,
+      id: params.sessionId,
+      timestamp: new Date().toISOString(),
+      cwd: process.cwd(),
+    };
+    fs.writeFileSync(params.transcriptPath, `${JSON.stringify(header)}\n`, "utf-8");
+    return { ok: true };
+  } catch (err) {
+    return { ok: false, error: err instanceof Error ? err.message : String(err) };
+  }
+}
+
+function appendAssistantTranscriptMessage(params: {
+  message: string;
+  label?: string;
+  sessionId: string;
+  storePath: string | undefined;
+  sessionFile?: string;
+  createIfMissing?: boolean;
+}): TranscriptAppendResult {
+  const transcriptPath = resolveTranscriptPath({
+    sessionId: params.sessionId,
+    storePath: params.storePath,
+    sessionFile: params.sessionFile,
+  });
+  if (!transcriptPath) {
+    return { ok: false, error: "transcript path not resolved" };
+  }
+
+  if (!fs.existsSync(transcriptPath)) {
+    if (!params.createIfMissing) {
+      return { ok: false, error: "transcript file not found" };
+    }
+    const ensured = ensureTranscriptFile({
+      transcriptPath,
+      sessionId: params.sessionId,
+    });
+    if (!ensured.ok) {
+      return { ok: false, error: ensured.error ?? "failed to create transcript file" };
+    }
+  }
+
+  const now = Date.now();
+  const messageId = randomUUID().slice(0, 8);
+  const labelPrefix = params.label ? `[${params.label}]\n\n` : "";
+  const messageBody: Record<string, unknown> = {
+    role: "assistant",
+    content: [{ type: "text", text: `${labelPrefix}${params.message}` }],
+    timestamp: now,
+    stopReason: "injected",
+    usage: { input: 0, output: 0, totalTokens: 0 },
+  };
+  const transcriptEntry = {
+    type: "message",
+    id: messageId,
+    timestamp: new Date(now).toISOString(),
+    message: messageBody,
+  };
+
+  try {
+    fs.appendFileSync(transcriptPath, `${JSON.stringify(transcriptEntry)}\n`, "utf-8");
+  } catch (err) {
+    return { ok: false, error: err instanceof Error ? err.message : String(err) };
+  }
+
+  return { ok: true, messageId, message: transcriptEntry.message };
+}
+
+function nextChatSeq(context: { agentRunSeq: Map<string, number> }, runId: string) {
+  const next = (context.agentRunSeq.get(runId) ?? 0) + 1;
+  context.agentRunSeq.set(runId, next);
+  return next;
+}
+
+function broadcastChatFinal(params: {
+  context: Pick<GatewayRequestContext, "broadcast" | "nodeSendToSession" | "agentRunSeq">;
+  runId: string;
+  sessionKey: string;
+  message?: Record<string, unknown>;
+}) {
+  const seq = nextChatSeq({ agentRunSeq: params.context.agentRunSeq }, params.runId);
+  const payload = {
+    runId: params.runId,
+    sessionKey: params.sessionKey,
+    seq,
+    state: "final" as const,
+    message: params.message,
+  };
+  params.context.broadcast("chat", payload);
+  params.context.nodeSendToSession(params.sessionKey, "chat", payload);
+}
+
+function broadcastChatError(params: {
+  context: Pick<GatewayRequestContext, "broadcast" | "nodeSendToSession" | "agentRunSeq">;
+  runId: string;
+  sessionKey: string;
+  errorMessage?: string;
+}) {
+  const seq = nextChatSeq({ agentRunSeq: params.context.agentRunSeq }, params.runId);
+  const payload = {
+    runId: params.runId,
+    sessionKey: params.sessionKey,
+    seq,
+    state: "error" as const,
+    errorMessage: params.errorMessage,
+  };
+  params.context.broadcast("chat", payload);
+  params.context.nodeSendToSession(params.sessionKey, "chat", payload);
+}
 
 export const chatHandlers: GatewayRequestHandlers = {
   "chat.history": async ({ params, respond, context }) => {
@@ -168,7 +293,7 @@ export const chatHandlers: GatewayRequestHandlers = {
       runIds: res.aborted ? [runId] : [],
     });
   },
-  "chat.send": async ({ params, respond, context }) => {
+  "chat.send": async ({ params, respond, context, client }) => {
     if (!validateChatSendParams(params)) {
       respond(
         false,
@@ -228,20 +353,13 @@ export const chatHandlers: GatewayRequestHandlers = {
         return;
       }
     }
-    const { cfg, storePath, entry, canonicalKey, store } = loadSessionEntry(p.sessionKey);
+    const { cfg, entry } = loadSessionEntry(p.sessionKey);
     const timeoutMs = resolveAgentTimeoutMs({
       cfg,
       overrideMs: p.timeoutMs,
     });
     const now = Date.now();
-    const sessionId = entry?.sessionId ?? randomUUID();
-    const sessionEntry = mergeSessionEntry(entry, {
-      sessionId,
-      updatedAt: now,
-    });
-    store[canonicalKey] = sessionEntry;
     const clientRunId = p.idempotencyKey;
-    registerAgentRunContext(clientRunId, { sessionKey: p.sessionKey });
 
     const sendPolicy = resolveSendPolicy({
       cfg,
@@ -298,21 +416,11 @@ export const chatHandlers: GatewayRequestHandlers = {
       const abortController = new AbortController();
       context.chatAbortControllers.set(clientRunId, {
         controller: abortController,
-        sessionId,
+        sessionId: entry?.sessionId ?? clientRunId,
         sessionKey: p.sessionKey,
         startedAtMs: now,
         expiresAtMs: resolveChatRunExpiresAtMs({ now, timeoutMs }),
       });
-      context.addChatRun(clientRunId, {
-        sessionKey: p.sessionKey,
-        clientRunId,
-      });
-
-      if (storePath) {
-        await updateSessionStore(storePath, (store) => {
-          store[canonicalKey] = sessionEntry;
-        });
-      }
 
       const ackPayload = {
         runId: clientRunId,
@@ -320,170 +428,116 @@ export const chatHandlers: GatewayRequestHandlers = {
       };
       respond(true, ackPayload, undefined, { runId: clientRunId });
 
-      if (isControlCommandMessage(parsedMessage, cfg)) {
-        try {
-          const isFastTestEnv = process.env.CLAWDBOT_TEST_FAST === "1";
-          const agentId = resolveSessionAgentId({ sessionKey: p.sessionKey, config: cfg });
-          const agentCfg = cfg.agents?.defaults;
-          const workspaceDir = resolveAgentWorkspaceDir(cfg, agentId);
-          const workspace = await ensureAgentWorkspace({
-            dir: workspaceDir,
-            ensureBootstrapFiles: !agentCfg?.skipBootstrap && !isFastTestEnv,
-          });
-          const ctx: MsgContext = {
-            Body: parsedMessage,
-            CommandBody: parsedMessage,
-            BodyForCommands: parsedMessage,
-            CommandSource: "text",
-            CommandAuthorized: true,
-            Provider: INTERNAL_MESSAGE_CHANNEL,
-            Surface: "tui",
-            From: p.sessionKey,
-            To: INTERNAL_MESSAGE_CHANNEL,
-            SessionKey: p.sessionKey,
-            ChatType: "direct",
-          };
-          const command = buildCommandContext({
-            ctx,
-            cfg,
-            agentId,
-            sessionKey: p.sessionKey,
-            isGroup: false,
-            triggerBodyNormalized: normalizeCommandBody(parsedMessage),
-            commandAuthorized: true,
-          });
-          const directives = parseInlineDirectives(parsedMessage);
-          const { provider, model } = resolveSessionModelRef(cfg, sessionEntry);
-          const contextTokens = resolveContextTokens({ agentCfg, model });
-          const resolveDefaultThinkingLevel = async () => {
-            const configured = agentCfg?.thinkingDefault;
-            if (configured) return configured;
-            const catalog = await context.loadGatewayModelCatalog();
-            return resolveThinkingDefault({ cfg, provider, model, catalog });
-          };
-          const resolvedThinkLevel =
-            normalizeThinkLevel(sessionEntry?.thinkingLevel ?? agentCfg?.thinkingDefault) ??
-            (await resolveDefaultThinkingLevel());
-          const resolvedVerboseLevel =
-            normalizeVerboseLevel(sessionEntry?.verboseLevel ?? agentCfg?.verboseDefault) ?? "off";
-          const resolvedReasoningLevel =
-            normalizeReasoningLevel(sessionEntry?.reasoningLevel) ?? "off";
-          const resolvedElevatedLevel = normalizeElevatedLevel(
-            sessionEntry?.elevatedLevel ?? agentCfg?.elevatedDefault,
-          );
-          const elevated = resolveElevatedPermissions({
-            cfg,
-            agentId,
-            ctx,
-            provider: INTERNAL_MESSAGE_CHANNEL,
-          });
-          const commandResult = await handleCommands({
-            ctx,
-            cfg,
-            command,
-            agentId,
-            directives,
-            elevated,
-            sessionEntry,
-            previousSessionEntry: entry,
-            sessionStore: store,
-            sessionKey: p.sessionKey,
-            storePath,
-            sessionScope: (cfg.session?.scope ?? "per-sender") as "per-sender" | "global",
-            workspaceDir: workspace.dir,
-            defaultGroupActivation: () => defaultGroupActivation(true),
-            resolvedThinkLevel,
-            resolvedVerboseLevel,
-            resolvedReasoningLevel,
-            resolvedElevatedLevel,
-            resolveDefaultThinkingLevel,
-            provider,
-            model,
-            contextTokens,
-            isGroup: false,
-          });
-          if (!commandResult.shouldContinue) {
-            const text = commandResult.reply?.text ?? "";
-            const message = {
-              role: "assistant",
-              content: text.trim() ? [{ type: "text", text }] : [],
-              timestamp: Date.now(),
-              command: true,
-            };
-            const payload = {
+      const trimmedMessage = parsedMessage.trim();
+      const injectThinking = Boolean(
+        p.thinking && trimmedMessage && !trimmedMessage.startsWith("/"),
+      );
+      const commandBody = injectThinking ? `/think ${p.thinking} ${parsedMessage}` : parsedMessage;
+      const clientInfo = client?.connect?.client;
+      const ctx: MsgContext = {
+        Body: parsedMessage,
+        BodyForAgent: parsedMessage,
+        BodyForCommands: commandBody,
+        RawBody: parsedMessage,
+        CommandBody: commandBody,
+        SessionKey: p.sessionKey,
+        Provider: INTERNAL_MESSAGE_CHANNEL,
+        Surface: INTERNAL_MESSAGE_CHANNEL,
+        OriginatingChannel: INTERNAL_MESSAGE_CHANNEL,
+        ChatType: "direct",
+        CommandAuthorized: true,
+        MessageSid: clientRunId,
+        SenderId: clientInfo?.id,
+        SenderName: clientInfo?.displayName,
+        SenderUsername: clientInfo?.displayName,
+      };
+
+      const agentId = resolveSessionAgentId({
+        sessionKey: p.sessionKey,
+        config: cfg,
+      });
+      let prefixContext: ResponsePrefixContext = {
+        identityName: resolveIdentityName(cfg, agentId),
+      };
+      const finalReplyParts: string[] = [];
+      const dispatcher = createReplyDispatcher({
+        responsePrefix: resolveEffectiveMessagesConfig(cfg, agentId).responsePrefix,
+        responsePrefixContextProvider: () => prefixContext,
+        onError: (err) => {
+          context.logGateway.warn(`webchat dispatch failed: ${formatForLog(err)}`);
+        },
+        deliver: async (payload, info) => {
+          if (info.kind !== "final") return;
+          const text = payload.text?.trim() ?? "";
+          if (!text) return;
+          finalReplyParts.push(text);
+        },
+      });
+
+      let agentRunStarted = false;
+      void dispatchInboundMessage({
+        ctx,
+        cfg,
+        dispatcher,
+        replyOptions: {
+          runId: clientRunId,
+          abortSignal: abortController.signal,
+          images: parsedImages.length > 0 ? parsedImages : undefined,
+          disableBlockStreaming: true,
+          onAgentRunStart: () => {
+            agentRunStarted = true;
+          },
+          onModelSelected: (ctx) => {
+            prefixContext.provider = ctx.provider;
+            prefixContext.model = extractShortModelName(ctx.model);
+            prefixContext.modelFull = `${ctx.provider}/${ctx.model}`;
+            prefixContext.thinkingLevel = ctx.thinkLevel ?? "off";
+          },
+        },
+      })
+        .then(() => {
+          if (!agentRunStarted) {
+            const combinedReply = finalReplyParts
+              .map((part) => part.trim())
+              .filter(Boolean)
+              .join("\n\n")
+              .trim();
+            let message: Record<string, unknown> | undefined;
+            if (combinedReply) {
+              const { storePath: latestStorePath, entry: latestEntry } = loadSessionEntry(
+                p.sessionKey,
+              );
+              const sessionId = latestEntry?.sessionId ?? entry?.sessionId ?? clientRunId;
+              const appended = appendAssistantTranscriptMessage({
+                message: combinedReply,
+                sessionId,
+                storePath: latestStorePath,
+                sessionFile: latestEntry?.sessionFile,
+                createIfMissing: true,
+              });
+              if (appended.ok) {
+                message = appended.message;
+              } else {
+                context.logGateway.warn(
+                  `webchat transcript append failed: ${appended.error ?? "unknown error"}`,
+                );
+                const now = Date.now();
+                message = {
+                  role: "assistant",
+                  content: [{ type: "text", text: combinedReply }],
+                  timestamp: now,
+                  stopReason: "injected",
+                  usage: { input: 0, output: 0, totalTokens: 0 },
+                };
+              }
+            }
+            broadcastChatFinal({
+              context,
               runId: clientRunId,
               sessionKey: p.sessionKey,
-              seq: 0,
-              state: "final" as const,
               message,
-            };
-            context.broadcast("chat", payload);
-            context.nodeSendToSession(p.sessionKey, "chat", payload);
-            context.dedupe.set(`chat:${clientRunId}`, {
-              ts: Date.now(),
-              ok: true,
-              payload: { runId: clientRunId, status: "ok" as const },
             });
-            context.chatAbortControllers.delete(clientRunId);
-            context.removeChatRun(clientRunId, clientRunId, p.sessionKey);
-            return;
           }
-        } catch (err) {
-          const payload = {
-            runId: clientRunId,
-            sessionKey: p.sessionKey,
-            seq: 0,
-            state: "error" as const,
-            errorMessage: formatForLog(err),
-          };
-          const error = errorShape(ErrorCodes.UNAVAILABLE, String(err));
-          context.broadcast("chat", payload);
-          context.nodeSendToSession(p.sessionKey, "chat", payload);
-          context.dedupe.set(`chat:${clientRunId}`, {
-            ts: Date.now(),
-            ok: false,
-            payload: {
-              runId: clientRunId,
-              status: "error" as const,
-              summary: String(err),
-            },
-            error,
-          });
-          context.chatAbortControllers.delete(clientRunId);
-          context.removeChatRun(clientRunId, clientRunId, p.sessionKey);
-          return;
-        }
-      }
-
-      const envelopeOptions = resolveEnvelopeFormatOptions(cfg);
-      const envelopedMessage = formatInboundEnvelope({
-        channel: "WebChat",
-        from: p.sessionKey,
-        timestamp: now,
-        body: parsedMessage,
-        chatType: "direct",
-        previousTimestamp: entry?.updatedAt,
-        envelope: envelopeOptions,
-      });
-      const lane = isAcpSessionKey(p.sessionKey) ? p.sessionKey : undefined;
-      void agentCommand(
-        {
-          message: envelopedMessage,
-          images: parsedImages.length > 0 ? parsedImages : undefined,
-          sessionId,
-          sessionKey: p.sessionKey,
-          runId: clientRunId,
-          thinking: p.thinking,
-          deliver: p.deliver,
-          timeout: Math.ceil(timeoutMs / 1000).toString(),
-          messageChannel: INTERNAL_MESSAGE_CHANNEL,
-          abortSignal: abortController.signal,
-          lane,
-        },
-        defaultRuntime,
-        context.deps,
-      )
-        .then(() => {
           context.dedupe.set(`chat:${clientRunId}`, {
             ts: Date.now(),
             ok: true,
@@ -502,6 +556,12 @@ export const chatHandlers: GatewayRequestHandlers = {
             },
             error,
           });
+          broadcastChatError({
+            context,
+            runId: clientRunId,
+            sessionKey: p.sessionKey,
+            errorMessage: String(err),
+          });
         })
         .finally(() => {
           context.chatAbortControllers.delete(clientRunId);