openclaw-scheduler 0.2.13 → 0.2.14

package/AGENTS.md

@@ -35,7 +35,7 @@ Check-in messages are delivered asynchronously. By the time they appear, the job
 
 ```bash
 # For chilisaus-dispatched jobs:
-node ~/.openclaw/chilisaus/index.mjs status --label <label>
+node ~/.openclaw/worktrees/openclaw-scheduler/dispatch/chilisaus.mjs status --label <label>
 
 # For scheduler jobs:
 openclaw-scheduler runs list <job-id> --json
@@ -321,3 +321,38 @@ See the
 for agentcli-specific agent instructions and the
 [MANIFEST-QUICK-REF.md](https://github.com/amittell/agentcli/blob/main/MANIFEST-QUICK-REF.md)
 for copy-paste manifest patterns.
+
+<!-- coord:begin -->
+<!-- AUTO-GENERATED by 'coord upgrade'. Do not hand-edit; next upgrade will overwrite. -->
+## Coordination protocol (mandatory)
+
+Multi-agent file coordination via the `coord` MCP server. Required calls every session:
+
+1. `list_claims` at task start.
+2. `claim_files` before editing.
+3. `release_claims` (or `release_session` to drop everything this MCP session holds, v0.6+) when done.
+
+If `claim_files` returns conflicts, stop and ask the user. No edits outside claimed scope; no opportunistic refactors; shared config edits only with explicit user approval.
+
+### Sub-file (symbol-level) claims (v0.14+)
+
+When two agents need different parts of the same file, claim symbols instead of whole files. Pass `symbols={"src/auth/login.ts": ["handleLogin"]}` to `claim_files` so the claim scopes to just that function. Two agents on disjoint symbols of the same file auto-coexist with no 409.
+
+- `Foo::handleA` (v0.16): method-level notation. Sibling methods of the same class auto-coexist; the bare class blocks all of its methods.
+- `Outer::Inner::method` (v0.17): recursive nesting works to any depth.
+- Server-side validation (v0.17): when `COORD_REPO_ROOT` is set, the server rejects symbols that do not exist in the file with a hint listing what is parseable. The MCP wrapper pre-validates locally before POST so typos fail fast; disable with `COORD_DISABLE_CLIENT_VALIDATION=1`.
+
+### Queueing instead of bouncing (v0.21+)
+
+When a hot file is contested and your work is blocking the team, pass `wait_seconds=60` to `claim_files`. The request joins a FIFO queue behind the blocking holder; on release the next queued requester is auto-granted. Pair with `urgency='low' | 'normal' | 'high' | 'blocking'` (v0.25) to jump ahead of lower-priority waiters; long-waiting entries age-boost one level after `COORD_QUEUE_AGE_BOOST_SECONDS` (v0.26). Abandon a wait early via `cancel_queue_request(queue_id, engineer=...)` (v0.26). Inspect your own queue rows via `my_requests(queued=True)` (v0.22).
+
+### Asking the holder directly (v0.6+ / v0.11+ decisions)
+
+If queueing will not work either:
+- `request_release` files an explicit ask against the holder's claim. The holder's TTL shortens; their decision lands back in your `my_requests` view.
+- `respond_to_request` decisions (v0.11+): `approved` (release whole claim), `denied` (keep it), `narrowed` (close the claim, open a tighter one -- pass `narrowed_pattern`), `coexist` (let the requester have a sibling claim on the same scope -- pass `coexist_pattern`). `coexist` is cooperative not enforced; agents on the same file still handle imports and module-level state themselves.
+
+### Operator visibility
+
+Poll `pending_requests` between operations to see who is blocked on your scope and respond via `respond_to_request`. The dashboard surfaces hotspot files, an auto-resolution heatmap, and a pending-queue panel for ambient awareness.
+<!-- coord:end -->

package/BEST-PRACTICES.md

@@ -368,7 +368,7 @@ Check-in messages (from `agent-checkin.mjs` or similar) are delivered asynchrono
 
 ```bash
 # Always do this before reporting job status:
-node ~/.openclaw/chilisaus/index.mjs status --label <label>
+node ~/.openclaw/worktrees/openclaw-scheduler/dispatch/chilisaus.mjs status --label <label>
 ```
 
 The `status` output gives you the authoritative `status` field (`accepted` / `running` / `done` / `error`), the last `updatedAt` timestamp, and the final `summary`. Use that — not the most recent check-in message.

package/CHANGELOG.md

@@ -2,6 +2,17 @@
 
 All notable changes to this project will be documented in this file.
 
+## [0.2.14] -- 2026-06-28
+
+### Fixed
+- fix(dispatch): honor configured dispatch default models from wrapper config, `DISPATCH_DEFAULT_MODEL`, and OpenClaw agent defaults before falling back to the static default
+- fix(dispatch): verify interrupted sessions before marking watcher failures
+- fix(dispatch): prefer human completion summaries in delivery flows
+
+### Changed
+- chore: upgrade coord to v0.35.0
+- docs(dispatch): retire old chilisaus fork paths and document migrated dispatch configuration
+
 ## [0.2.13] -- 2026-06-24
 
 ### Fixed

package/README.md

@@ -1767,7 +1767,7 @@ For normal chat-triggered dispatches, always pass `--deliver-to` from the inboun
 | `--message-stdin` | -- | Read the prompt from stdin explicitly. If stdin is piped and no explicit prompt source is set, dispatch auto-reads stdin. |
 | `--mode` | `fresh` | `fresh` creates a new session. `reuse` continues the last session recorded for this label. |
 | `--thinking` | -- | Reasoning budget: `low`, `high`, or `xhigh`. |
-| `--model` | -- | Model override, e.g. `anthropic/claude-sonnet-4-6`. |
+| `--model` | configured dispatch default | Model override, e.g. `anthropic/claude-sonnet-4-6`. When omitted, dispatch uses wrapper `config.defaultModel`, wrapper `config.dispatch.model`, `DISPATCH_DEFAULT_MODEL`, `agents.defaults.dispatch.model`, `agents.defaults.model`, then the built-in fallback. |
 | `--deliver-to` | -- | Delivery target (e.g. Telegram chat ID). Registers the scheduler watcher job for durable final delivery. The gateway spawn itself stays fire-and-forget so raw tool output and internal done payloads cannot leak directly to chat. Chat-triggered callers should pass inbound metadata `chat_id` here, especially for group chats. |
 | `--delivery-mode` | `announce` | `announce` delivers only when output is non-empty. `announce-always` delivers unconditionally. `none` suppresses delivery. |
 | `--timeout` | `300` | Session timeout in seconds. |

package/dispatch/README.md

@@ -20,8 +20,9 @@ No scheduler DB dependency. No dispatcher tick delay. Sessions start instantly.
 | `watcher.mjs` | Delivery monitoring process |
 | `529-recovery.mjs` | Transient error recovery |
 | `deliver-watcher.sh` | Shell wrapper for result retrieval |
-| `chilisaus.mjs` | Branded wrapper |
+| `chilisaus.mjs` | Branded chilisaus wrapper over this dispatch engine |
 | `config.example.json` | Example config |
+| `chilisaus.config.example.json` | Chilisaus branding config example |
 | `test-done-postoffice.mjs` | Done handler test |
 | `~/.openclaw/scheduler/dispatch/labels.json` | Durable label→session ledger |
 | `README.md` | This file |
@@ -50,6 +51,20 @@ Orchestrator calls:
   → hooks.mjs fires dispatch.started to Loki
 ```
 
+### Chilisaus branding
+
+Chilisaus is not a separate implementation. It is the branded entrypoint in this
+directory:
+
+```bash
+node dispatch/chilisaus.mjs enqueue --label ticket-42 --message "Fix it"
+```
+
+`chilisaus.mjs` sets `DISPATCH_CONFIG_DIR` to this dispatch directory, then loads
+`index.mjs`. To use the historical chilisaus branding, copy
+`chilisaus.config.example.json` to `config.json` in the same directory. Runtime
+changes belong in this `openclaw-scheduler/dispatch` tree.
+
 ---
 
 ## Subcommands
@@ -86,7 +101,7 @@ cat prompt.md | node dispatch/index.mjs enqueue \
 | `--mode` | `fresh` | `fresh` = new session; `reuse` = continue last session for this label |
 | `--session-key` | — | Explicit session key (bypasses ledger lookup) |
 | `--agent` | `main` | Agent ID |
-| `--model` | — | Model override (e.g. `anthropic/claude-sonnet-4-6`) |
+| `--model` | configured dispatch default | Model override (e.g. `anthropic/claude-sonnet-4-6`). When omitted, dispatch uses wrapper `config.defaultModel`, wrapper `config.dispatch.model`, `DISPATCH_DEFAULT_MODEL`, `agents.defaults.dispatch.model`, `agents.defaults.model`, then the built-in fallback. |
 | `--thinking` | — | Reasoning level: `low`, `high`, `xhigh` |
 | `--timeout` | `300` | Seconds before run times out |
 | `--deliver-to` | — | Delivery target (chat ID, channel ID, handle, etc.). Enables `deliver:true` on the gateway call. Chat-triggered callers should pass inbound metadata `chat_id` here, especially for group chats. |

package/dispatch/chilisaus.config.example.json

@@ -0,0 +1,10 @@
+{
+  "_comment": "Chilisaus branding config. Copy to config.json next to dispatch/chilisaus.mjs for a branded deployment.",
+
+  "name": "chilisaus",
+  "brand": "chilisaus 🌶️",
+
+  "startupGraceMs": 90000,
+  "stuckThresholdMs": 600000,
+  "maxWatcherAgeMs": 7200000
+}

package/dispatch/completion.mjs

@@ -39,7 +39,6 @@ const HUMAN_SUMMARY_SECTION_RE = /(?:^|\n)\s*(?:Human-readable summary|Human sum
 const TECHNICAL_DETAILS_SECTION_RE = /(?:^|\n)\s*(?:Technical details?|Details(?:_technical)?)\s*:\s*/i;
 const HUMAN_SUMMARY_LABEL_RE = /^(?:human-readable summary|human summary)\s*:\s*/i;
 const TECHNICAL_DETAILS_LABEL_RE = /^(?:technical details?|details(?:_technical)?)\s*:\s*/i;
-const FORMATTER_META_SUMMARY_RE = /^Final completion updates now (?:start with a short plain-English summary|arrive as one clean plain-English summary)\./i;
 const FINAL_REPORT_HEADING_RE = /^(?:#{1,6}\s*)?(?:root cause|files? changed|changes|validation|tests?(?: run| passed)?|sacrificial(?: delivery)?(?: result)?|deployment(?:\/live-runtime)?(?: step)?|live-runtime(?: step)?|result|results|summary|highlights?|notes?|follow[- ]ups?|next steps?|blockers?|implementation|what changed|verification)\s*:?$/i;
 const FINAL_REPORT_CUE_RE = /\b(?:root cause|files? changed|tests? run|validation|sacrificial(?: delivery)?(?: result)?|deployment(?:\/live-runtime)?(?: step)?|live-runtime(?: step)?|final report|human-readable report|files changed|tests passed)\b/i;
 
@@ -752,42 +751,12 @@ function buildCompletionLeadFromThemes(themes) {
   return truncateText(sentences.join(' '), MAX_DELIVERY_CHARS);
 }
 
-function looksLikeFormatterMetaSummary(text) {
-  const normalized = normalizeCompletionText(text);
-  return Boolean(normalized && FORMATTER_META_SUMMARY_RE.test(normalized));
-}
-
 function getCompletionRawSummary(completion) {
   const details = getCompletionTechnicalDetails(completion);
   if (!details || typeof details !== 'object') return null;
   return normalizeCompletionText(details.raw_summary ?? details.rawSummary);
 }
 
-function looksLikeCompletionFormatterImplementation(text) {
-  const normalized = normalizeCompletionText(text);
-  if (!normalized) return false;
-
-  const fragments = extractTechnicalFragments(normalized);
-  const themes = detectTechnicalThemes(normalized, fragments);
-  if (!themes.completionFlow) return false;
-
-  return TECHNICAL_COMMIT_PREFIX_RE.test(cleanMarkdown(normalized).replace(/\s+/g, ' ').trim())
-    || /\b(?:completion|deliveryText|summary_human|summaryHuman|details_technical|resolveCompletionDelivery|buildTerminalCompletionPayload|payload-precedence|watcher path|completion watcher|final completion updates?)\b/i.test(normalized);
-}
-
-function getRawSummaryOverrideForFormatterMeta(completion, structuredTexts) {
-  if (!structuredTexts.some(looksLikeFormatterMetaSummary)) return null;
-
-  const rawSummary = getCompletionRawSummary(completion);
-  if (!rawSummary) return null;
-  if (isGenericOrTrivial(rawSummary)) return null;
-  if (isInternalTransportNoiseText(rawSummary)) return null;
-  if (looksLikeRawPayloadText(rawSummary)) return null;
-  if (looksLikeCompletionFormatterImplementation(rawSummary)) return null;
-
-  return truncateText(rawSummary, MAX_DELIVERY_CHARS);
-}
-
 function buildHumanizedTechnicalSummary(rawText, fallbackSummary) {
   const cleanedRaw = normalizeCompletionText(rawText);
   const fallback = normalizeCompletionText(fallbackSummary);
@@ -888,7 +857,7 @@ function summarizeChecklistTechnicalDetails(checklist, sha) {
   return parts.length > 0 ? `Checks: ${parts.join('; ')}.` : null;
 }
 
-function buildTechnicalDetailsText({ rawText, summaryText, completion } = {}) {
+function buildTechnicalDetailsText({ rawText, summaryText, completion, includeRawSummaryDetails = true } = {}) {
   const raw = normalizeCompletionText(rawText);
   const summary = normalizeCompletionText(summaryText);
   const details = getCompletionTechnicalDetails(completion);
@@ -922,7 +891,7 @@ function buildTechnicalDetailsText({ rawText, summaryText, completion } = {}) {
       && (!rawTechnical || normalized !== rawTechnicalSource)) {
       parts.push(truncateText(normalized, 220));
     }
-  } else if (details && typeof details === 'object') {
+  } else if (includeRawSummaryDetails && details && typeof details === 'object') {
     const rawSummary = normalizeCompletionText(details.raw_summary);
     const detailSummarySections = extractStructuredSummarySections(rawSummary);
     const splitDetailSummary = extractExplicitTechnicalTail(rawSummary);
@@ -1019,7 +988,7 @@ export function isMeaningfulCompletionText(value) {
 }
 
 function getCompletionSummaryHuman(completion) {
-  return normalizeCompletionText(completion?.summary_human ?? completion?.summaryHuman);
+  return normalizeCompletionText(completion?.summary_human ?? completion?.summaryHuman ?? completion?.prose);
 }
 
 function getCompletionTechnicalDetails(completion) {
@@ -1328,12 +1297,14 @@ export function resolveCompletionDelivery({ lastReply, completion, fallbackSumma
   const rawCompletionSummaryHuman = getCompletionSummaryHuman(completion);
   const rawCompletionSummary = normalizeCompletionText(completion?.summary);
   const rawCompletionDelivery = normalizeCompletionText(completion?.deliveryText);
+  const rawCompletionRawSummary = getCompletionRawSummary(completion);
   const rawFallback = normalizeCompletionText(fallbackSummary);
 
   const reply = humanizeCompletionText(lastReply);
   const completionSummaryHuman = humanizeCompletionText(rawCompletionSummaryHuman);
   const completionSummary = humanizeCompletionText(completion?.summary);
   const completionDelivery = humanizeCompletionText(completion?.deliveryText);
+  const completionRawSummary = humanizeCompletionText(rawCompletionRawSummary);
   const fallback = humanizeCompletionText(fallbackSummary);
   const synthesizedFromTechnical = synthesizeCompletionReply({
     checklist: completion?.checklist,
@@ -1344,23 +1315,12 @@ export function resolveCompletionDelivery({ lastReply, completion, fallbackSumma
   const authoritativeStructuredSummary = completionDeliverySource && completionDeliverySource !== 'technical-synthesis'
     ? preferredSummary
     : null;
-  const rawSummaryOverride = getRawSummaryOverrideForFormatterMeta(completion, [
-    rawCompletionSummaryHuman,
-    rawCompletionSummary,
-    rawCompletionDelivery,
-  ]);
-  if (rawSummaryOverride) {
-    return {
-      deliveryText: rawSummaryOverride,
-      summary: rawSummaryOverride,
-      source: 'raw-summary',
-    };
-  }
   const noisyTexts = [
     rawReply,
     rawCompletionSummaryHuman,
     rawCompletionDelivery,
     rawCompletionSummary,
+    rawCompletionRawSummary,
     rawFallback,
   ].filter(isInternalTransportNoiseText);
   const isDeliverableText = (rawText, summarizedText) => Boolean(summarizedText)
@@ -1395,6 +1355,7 @@ export function resolveCompletionDelivery({ lastReply, completion, fallbackSumma
       rawText: candidate.rawText,
       summaryText: candidate.text,
       completion,
+      includeRawSummaryDetails: false,
     });
     return {
       deliveryText: composeDeliveryText(candidate.text, technicalDetailsText),
@@ -1408,6 +1369,7 @@ export function resolveCompletionDelivery({ lastReply, completion, fallbackSumma
       rawText: rawReply,
       summaryText: reply,
       completion,
+      includeRawSummaryDetails: false,
     });
     return {
       deliveryText: composeDeliveryText(reply, technicalDetailsText),
@@ -1416,6 +1378,19 @@ export function resolveCompletionDelivery({ lastReply, completion, fallbackSumma
     };
   }
 
+  if (isDeliverableText(rawCompletionRawSummary, completionRawSummary)) {
+    const technicalDetailsText = buildTechnicalDetailsText({
+      rawText: rawCompletionRawSummary,
+      summaryText: completionRawSummary,
+      completion,
+    });
+    return {
+      deliveryText: composeDeliveryText(completionRawSummary, technicalDetailsText),
+      summary: completionRawSummary,
+      source: 'raw-summary',
+    };
+  }
+
   if (isDeliverableText(synthesizedFromTechnical, synthesizedFromTechnical)) {
     return {
       deliveryText: synthesizedFromTechnical,

package/dispatch/default-model.mjs

@@ -0,0 +1,83 @@
+import { existsSync, readFileSync } from 'fs';
+import { join, resolve as pathResolve } from 'path';
+
+export const STATIC_DISPATCH_DEFAULT_MODEL = 'openai/gpt-5.5';
+
+function isRecord(value) {
+  return Boolean(value) && typeof value === 'object' && !Array.isArray(value);
+}
+
+function normalizeString(value) {
+  return typeof value === 'string' && value.trim() ? value.trim() : null;
+}
+
+function resolveModelPrimary(value) {
+  const direct = normalizeString(value);
+  if (direct) return direct;
+  if (!isRecord(value)) return null;
+  return normalizeString(value.primary);
+}
+
+function resolveHomeRelativePath(input, homeDir) {
+  const trimmed = normalizeString(input);
+  if (!trimmed) return null;
+  if (trimmed === '~') return homeDir;
+  if (trimmed.startsWith('~/')) return join(homeDir, trimmed.slice(2));
+  return pathResolve(trimmed);
+}
+
+function resolveOpenClawStateDir({ env, homeDir }) {
+  return resolveHomeRelativePath(env.OPENCLAW_STATE_DIR, homeDir) || join(homeDir, '.openclaw');
+}
+
+export function resolveOpenClawConfigPath({ env = process.env, homeDir } = {}) {
+  const effectiveHome = homeDir || env.HOME || process.env.HOME;
+  if (!effectiveHome) return null;
+  return (
+    resolveHomeRelativePath(env.OPENCLAW_CONFIG_PATH, effectiveHome) ||
+    join(resolveOpenClawStateDir({ env, homeDir: effectiveHome }), 'openclaw.json')
+  );
+}
+
+export function readOpenClawConfig({
+  env = process.env,
+  homeDir,
+  exists = existsSync,
+  readFile = readFileSync,
+} = {}) {
+  const configPath = resolveOpenClawConfigPath({ env, homeDir });
+  if (!configPath || !exists(configPath)) return {};
+  try {
+    return JSON.parse(readFile(configPath, 'utf8'));
+  } catch {
+    return {};
+  }
+}
+
+export function resolveOpenClawDispatchDefaultModel(config) {
+  const defaults = isRecord(config?.agents?.defaults) ? config.agents.defaults : {};
+  return (
+    resolveModelPrimary(defaults.dispatch?.model) ||
+    resolveModelPrimary(defaults.model) ||
+    null
+  );
+}
+
+export function resolveDefaultDispatchModel({
+  dispatchConfig = {},
+  openClawConfig,
+  env = process.env,
+  homeDir,
+  exists = existsSync,
+  readFile = readFileSync,
+} = {}) {
+  return (
+    normalizeString(dispatchConfig.defaultModel) ||
+    resolveModelPrimary(dispatchConfig.dispatch?.model) ||
+    normalizeString(env.DISPATCH_DEFAULT_MODEL) ||
+    resolveOpenClawDispatchDefaultModel(
+      openClawConfig ?? readOpenClawConfig({ env, homeDir, exists, readFile }),
+    ) ||
+    STATIC_DISPATCH_DEFAULT_MODEL
+  );
+}

package/dispatch/index.mjs

@@ -45,6 +45,7 @@ import { getDispatchLivenessPolicy } from './liveness.mjs';
 import { resolveLabelsPath } from './paths.mjs';
 import { onStarted, onFinished, onStuck } from './hooks.mjs';
 import { resolveMessageInput } from './message-input.mjs';
+import { resolveDefaultDispatchModel } from './default-model.mjs';
 import { buildDispatchDeliverySurface } from '../scripts/dispatch-cli-utils.mjs';
 
 const __dirname = dirname(fileURLToPath(import.meta.url));
@@ -96,6 +97,11 @@ function loadConfig() {
 
 const config = loadConfig();
 const BRAND = config.name ?? 'dispatch';
+const DEFAULT_DISPATCH_MODEL = resolveDefaultDispatchModel({
+  dispatchConfig: config,
+  env: process.env,
+  homeDir: HOME_DIR,
+});
 
 /** Load gateway auth token from config or env */
 function getGatewayToken() {
@@ -1096,7 +1102,7 @@ async function cmdEnqueue(flags) {
 
   // Dynamic branding: resolve per-agent brand name
   const agentBrand = config.agents?.[agent]?.name || (agent !== 'main' ? agent : null) || config.name || 'dispatch';
-  const model       = flags.model            || null;
+  const model       = flags.model            || DEFAULT_DISPATCH_MODEL;
 
   // -- Session key resolution ----------------------------------
   let sessionKey = flags['session-key'] || null;

package/dispatch/watcher.mjs

@@ -927,6 +927,32 @@ function markLabelError(label, errorSummary) {
   }
 }
 
+function getLabelEntry(label) {
+  try {
+    const labels = loadLabels();
+    return labels[label] || null;
+  } catch (err) {
+    process.stderr.write(`[watcher] label load failed for ${label}: ${err.message}\n`);
+    return null;
+  }
+}
+
+function runVerifyCmd(label, entry) {
+  if (!entry?.verifyCmd) return { configured: false, ok: false };
+
+  process.stderr.write(`[watcher] running verify-cmd for ${label}: ${entry.verifyCmd}\n`);
+  try {
+    execSync(entry.verifyCmd, { stdio: 'pipe', timeout: 60000, shell: true });
+    process.stderr.write(`[watcher] verify-cmd passed for ${label}\n`);
+    return { configured: true, ok: true };
+  } catch (verifyErr) {
+    const stderr = verifyErr.stderr ? verifyErr.stderr.toString().trim() : verifyErr.message;
+    const message = stderr || `exit code ${verifyErr.status ?? 1}`;
+    process.stderr.write(`[watcher] verify-cmd failed: ${message}\n`);
+    return { configured: true, ok: false, message };
+  }
+}
+
 let exitZeroOnTerminal = false;
 
 /**
@@ -941,33 +967,20 @@ function deliverResult(label, lastReply, fallbackSummary, completionPayload = nu
   // -- verify-cmd check -----------------------------------------------------
   // Run the stored verify-cmd (if any) before declaring the job done.
   // A non-zero exit flips the job to error state and sends an alert instead.
-  try {
-    const labels = loadLabels();
-    const entry = labels[label];
-    if (entry?.verifyCmd) {
-      process.stderr.write(`[watcher] running verify-cmd for ${label}: ${entry.verifyCmd}\n`);
-      try {
-        execSync(entry.verifyCmd, { stdio: 'pipe', timeout: 60000, shell: true });
-        process.stderr.write(`[watcher] verify-cmd passed for ${label}\n`);
-      } catch (verifyErr) {
-        const stderr = verifyErr.stderr ? verifyErr.stderr.toString().trim() : verifyErr.message;
-        const errMsg = `verify-cmd failed: ${stderr || 'exit code ' + (verifyErr.status ?? 1)}`;
-        process.stderr.write(`[watcher] ${errMsg}\n`);
-        markLabelError(label, errMsg);
-        // Output failure notice -- scheduler delivers this to the delivery target
-        process.stdout.write(
-          `🌶️ *dispatch* [${label}] ⚠️ VERIFICATION FAILED\n\n` +
-          `The agent session completed but the post-completion verify-cmd exited non-zero.\n\n` +
-          `**Verify command:** \`${entry.verifyCmd}\`\n` +
-          `**Error:** ${stderr || 'non-zero exit'}\n\n` +
-          `Job marked as \`error\`. The agent may have reported done without completing the actual work.\n`
-        );
-        process.exit(exitZeroOnTerminal ? 0 : 1);
-      }
-    }
-  } catch (loadErr) {
-    // Non-fatal -- if labels can't be read, skip verify check and proceed normally
-    process.stderr.write(`[watcher] verify-cmd check skipped (labels load error): ${loadErr.message}\n`);
+  const entry = getLabelEntry(label);
+  const verify = runVerifyCmd(label, entry);
+  if (verify.configured && !verify.ok) {
+    const errMsg = `verify-cmd failed: ${verify.message || 'non-zero exit'}`;
+    markLabelError(label, errMsg);
+    // Output failure notice -- scheduler delivers this to the delivery target
+    process.stdout.write(
+      `🌶️ *dispatch* [${label}] ⚠️ VERIFICATION FAILED\n\n` +
+      `The agent session completed but the post-completion verify-cmd exited non-zero.\n\n` +
+      `**Verify command:** \`${entry.verifyCmd}\`\n` +
+      `**Error:** ${verify.message || 'non-zero exit'}\n\n` +
+      `Job marked as \`error\`. The agent may have reported done without completing the actual work.\n`
+    );
+    process.exit(exitZeroOnTerminal ? 0 : 1);
   }
 
   // Update labels.json before exiting -- prevents stuck detector false positives
@@ -995,6 +1008,27 @@ function deliverResult(label, lastReply, fallbackSummary, completionPayload = nu
 
 function emitInterruptedOutcome(label, summary, result = null) {
   process.stderr.write(`[watcher] [${label}] session auto-resolved as interrupted -- work may be incomplete\n`);
+  const entry = getLabelEntry(label);
+  const verify = runVerifyCmd(label, entry);
+  if (verify.configured && verify.ok) {
+    const recoveredReply = result?.lastReply || result?.diagnosticReply || null;
+    const completionPayload = result?.completion || (recoveredReply ? null : {
+      summary_human: summary
+        ? `The session stopped before sending the done signal, but verification passed. ${summary}`
+        : 'The session stopped before sending the done signal, but verification passed.',
+      details: {
+        interrupted: true,
+        verify_cmd: entry.verifyCmd,
+      },
+    });
+    deliverResult(
+      label,
+      recoveredReply,
+      'completed after interrupted session; verify-cmd passed',
+      completionPayload,
+    );
+  }
+
   markLabelError(label, summary || 'interrupted: session went idle without calling done');
   process.stdout.write(
     `⚠️ dispatch [${label}] session went idle before completing -- work may be incomplete` +

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "openclaw-scheduler",
-  "version": "0.2.13",
+  "version": "0.2.14",
   "description": "SQLite-backed job scheduler and workflow engine for OpenClaw agents",
   "type": "module",
   "main": "./index.js",
@@ -17,7 +17,7 @@
     "./package.json": "./package.json"
   },
   "engines": {
-    "node": "22.x || 24.x || 26.x"
+    "node": "22.x || 24.x || 25.x || 26.x"
   },
   "scripts": {
     "start": "node dispatcher.js",
@@ -39,6 +39,8 @@
     "dispatch/529-recovery.mjs",
     "dispatch/completion.mjs",
     "dispatch/config.example.json",
+    "dispatch/chilisaus.config.example.json",
+    "dispatch/default-model.mjs",
     "dispatch/deliver-watcher.sh",
     "dispatch/hooks.mjs",
     "dispatch/index.mjs",