@yemi33/minions 0.1.1610 → 0.1.1612

package/CHANGELOG.md

@@ -1,6 +1,6 @@
 # Changelog
 
-## 0.1.1610 (2026-04-28)
+## 0.1.1612 (2026-04-28)
 
 ### Features
 - hard-pin agent assignment when CC names a specific agent
@@ -9,6 +9,7 @@
 - replace runtime text tag with inline SVG logos
 
 ### Fixes
+-  suppress failure inbox notes
 -  kill agents from engine tmp pid files
 -  gate live completion banner on final process exit
 - wrap CC Model input so its label/hint survive runtime hydration

package/dashboard.js

@@ -5201,6 +5201,7 @@ What would you like to discuss or change? When you're happy, say "approve" and I
       engine: { state: engine.state, pid: engine.pid },
       agents: agents.map(a => ({ id: a.id, name: a.name, status: a.status })),
       projects: PROJECTS.map(p => ({ name: p.name, reachable: fs.existsSync(p.localPath) })),
+      minionsDir: MINIONS_DIR,
       uptime: process.uptime(),
       timestamp: new Date().toISOString()
     };

package/engine/dispatch.js

@@ -220,7 +220,7 @@ function completeDispatch(id, result = DISPATCH_RESULT.SUCCESS, reason = '', res
           ? `Non-retryable failure: ${effectiveReason}${classSuffix}`
           : (reason || `Failed after ${maxRetries} retries${classSuffix}`);
         lifecycle().updateWorkItemStatus(item.meta, WI_STATUS.FAILED, finalReason);
-        // Alert: find items blocked by this failure and write inbox note
+        // Surface blocked dependents in logs without creating failure inbox noise.
         try {
           const config = getConfig();
           const failedId = item.meta.item.id;
@@ -229,16 +229,9 @@ function completeDispatch(id, result = DISPATCH_RESULT.SUCCESS, reason = '', res
           allItems.filter(w => w.status === WI_STATUS.PENDING && (w.depends_on || []).includes(failedId))
             .forEach(w => blockedItems.push(`- \`${w.id}\` — ${w.title}`));
 
-          writeInboxAlert(`failed-${failedId}`,
-            `# Work Item Failed — \`${failedId}\`\n\n` +
-            `**Item:** ${item.meta.item.title || failedId}\n` +
-            `**Reason:** ${finalReason}\n\n` +
-            (blockedItems.length > 0
-              ? `**Blocked dependents (${blockedItems.length}):**\n${blockedItems.join('\n')}\n\n` +
-                `These items cannot dispatch until \`${failedId}\` is fixed and reset to \`pending\`.\n`
-              : `No downstream items are blocked.\n`)
-          );
-        } catch (e) { log('warn', 'write failure alert: ' + e.message); }
+          log('warn', `Work item ${failedId} failed: ${finalReason}` +
+            (blockedItems.length > 0 ? `; blocked dependents: ${blockedItems.map(line => line.replace(/^- `([^`]+)`.*/, '$1')).join(', ')}` : '; no downstream items blocked'));
+        } catch (e) { log('warn', 'summarize failure dependents: ' + e.message); }
       }
     }
 

package/engine/lifecycle.js

@@ -69,19 +69,11 @@ function checkPlanCompletion(meta, config) {
   const doneItems = planItems.filter(w => DONE_STATUSES.has(w.status));
   const failedItems = planItems.filter(w => w.status === WI_STATUS.FAILED || w.status === WI_STATUS.CANCELLED);
 
-  // Escalate failed/cancelled items to human — write inbox alert (deduped by slug)
   if (failedItems.length > 0) {
-    const alertSlug = `plan-failure-escalation-${planFile.replace('.json', '')}`;
     const failDetails = failedItems.map(w =>
       `- \`${w.id}\`: ${w.title || 'Unknown'} — ${w.failReason || w.status}`
     ).join('\n');
-    shared.writeToInbox('engine', alertSlug,
-      `# Plan Items Failed: ${plan.plan_summary || planFile}\n\n` +
-      `**${failedItems.length}** of ${planFeatureIds.size} item(s) failed or were cancelled:\n\n${failDetails}\n\n` +
-      `The plan is completing with partial results (${doneItems.length} done, ${failedItems.length} failed).\n` +
-      `Review failed items and re-dispatch manually if needed.\n`
-    );
-    log('warn', `Plan ${planFile}: ${failedItems.length} item(s) failed/cancelled — escalating to human: ${failedItems.map(w => w.id).join(', ')}`);
+    log('warn', `Plan ${planFile}: ${failedItems.length} item(s) failed/cancelled; completing with partial results (${doneItems.length} done, ${failedItems.length} failed):\n${failDetails}`);
   }
 
   // 1. Mark plan as completed
@@ -1127,8 +1119,7 @@ async function processPendingRebases(config) {
       if (entry.attempts < 3) {
         remaining.push(entry);
       } else {
-        shared.writeToInbox('engine', `rebase-fail-${pr.id}`,
-          `# Rebase Failed: ${pr.id}\n\nBranch \`${pr.branch}\` could not be rebased onto main after dependency ${entry.mergedItemId} merged.\n\nError: ${result.error}\n\nManual rebase may be needed.`);
+        log('warn', `Rebase failed after retries for ${pr.id} on ${pr.branch}: ${result.error}`);
       }
     }
   }
@@ -1238,8 +1229,7 @@ async function handlePostMerge(pr, project, config, newStatus) {
           }
           const result = await rebaseBranchOntoMain(depPr, depProject, config);
           if (!result.success) {
-            shared.writeToInbox('engine', `rebase-fail-${depPr.id}`,
-              `# Rebase Failed: ${depPr.id}\n\nBranch \`${depPr.branch}\` could not be rebased onto main after dependency ${mergedItemId} merged.\n\nError: ${result.error}\n\nManual rebase may be needed.`);
+            log('warn', `Rebase failed for ${depPr.id} on ${depPr.branch} after dependency ${mergedItemId} merged: ${result.error}`);
           }
         }
       }
@@ -1820,19 +1810,23 @@ async function runPostCompletionHooks(dispatchItem, agentId, code, stdout, confi
         };
         return runs;
       }, { defaultValue: {} });
-      // Write a completion note to inbox with back-references
-      const noteSlug = `sched-completion-${scheduleId}`;
       const status = effectiveSuccess ? 'succeeded' : 'failed';
-      const noteContent = `# Scheduled Task ${status}: ${meta.item.title || scheduleId}\n\n` +
-        `**Schedule:** \`${scheduleId}\`\n` +
-        `**Work Item:** \`${itemId}\`\n` +
-        `**Result:** ${status}\n` +
-        (resultSummary ? `\n## Summary\n${resultSummary}\n` : '');
-      shared.writeToInbox('engine', noteSlug, noteContent, null, {
-        sourceItem: itemId,
-        scheduleId,
-      });
-      log('info', `Scheduled task ${scheduleId} (${itemId}) → ${status}, back-reference written`);
+      if (effectiveSuccess) {
+        // Write a completion note to inbox with back-references only for successful scheduled runs.
+        const noteSlug = `sched-completion-${scheduleId}`;
+        const noteContent = `# Scheduled Task ${status}: ${meta.item.title || scheduleId}\n\n` +
+          `**Schedule:** \`${scheduleId}\`\n` +
+          `**Work Item:** \`${itemId}\`\n` +
+          `**Result:** ${status}\n` +
+          (resultSummary ? `\n## Summary\n${resultSummary}\n` : '');
+        shared.writeToInbox('engine', noteSlug, noteContent, null, {
+          sourceItem: itemId,
+          scheduleId,
+        });
+        log('info', `Scheduled task ${scheduleId} (${itemId}) → ${status}, back-reference written`);
+      } else {
+        log('warn', `Scheduled task ${scheduleId} (${itemId}) → ${status}; inbox note suppressed`);
+      }
     } catch (err) { log('warn', `Scheduled task back-reference: ${err.message}`); }
   }
 

package/engine.js

@@ -1030,13 +1030,6 @@ async function spawnAgent(dispatchItem, config) {
         _trustCheckDone = true;
         updateAgentStatus(id, AGENT_STATUS.TRUST_BLOCKED, 'Agent appears to be waiting for trust approval');
         log('warn', `Trust gate detected for ${agentId} (${id}) — agent may be blocked on a permission prompt`);
-        writeInboxAlert(`trust-blocked-${id}`,
-          `# Trust Gate Blocked — \`${id}\`\n\n` +
-          `**Agent:** ${agentId}\n` +
-          `**Task:** ${dispatchItem.task || type}\n\n` +
-          `The agent appears to be blocked on a trust/permission prompt within 30s of spawn.\n` +
-          `Check the agent's live output and approve the trust gate manually.\n`
-        );
       }
     } else if (!_trustCheckDone) {
       _trustCheckDone = true; // past 30s window
@@ -1867,12 +1860,6 @@ function materializePlansAsWorkItems(config) {
       if (cycles.length > 0) {
         log('error', `Dependency cycle detected in plan ${file}: ${cycles.join(', ')} — skipping cyclic items`);
         cycles.forEach(c => cycleSet.add(c));
-        writeInboxAlert(`cycle-${path.basename(file, '.json')}`,
-          `# Dependency Cycle Detected — ${path.basename(file)}\n\n` +
-          `The following PRD items form a cycle and were **skipped** (will never be dispatched):\n\n` +
-          cycles.map(id => `- \`${id}\``).join('\n') + '\n\n' +
-          `Fix by removing or reordering the \`depends_on\` relationships in \`prd/${path.basename(file)}\`.\n`
-        );
       }
     }
 
@@ -2087,10 +2074,6 @@ async function discoverFromPrs(config, project) {
     const evalCycles = pr._reviewFixCycles || 0;
     const evalEscalated = evalCycles >= evalMax;
     if (evalEscalated && !pr._evalEscalated) {
-      writeInboxAlert(`eval-escalated-${pr.agent || 'unassigned'}-${prDisplayId}`,
-        `# Review Loop Escalation\n\n**PR ${pr.id}**: ${pr.title || ''} on branch \`${pr.branch || 'unknown'}\` has gone through **${evalCycles}** review→fix cycles without approval.\n\n` +
-        `Last review: ${pr.minionsReview?.note ? pr.minionsReview.note.slice(0, 200) : 'See PR thread'}\n\n` +
-        `Auto-dispatch of reviews and fixes has been suspended. Please review the PR manually.`);
       try {
         mutatePullRequests(projectPrPath(project), prs => {
           const target = shared.findPrRecord(prs, pr, project);
@@ -2258,12 +2241,6 @@ async function discoverFromPrs(config, project) {
       // Check if max retry cap reached — escalate to human instead of dispatching another fix
       if ((pr.buildFixAttempts || 0) >= maxBuildFix) {
         if (!pr.buildFixEscalated) {
-          writeInboxAlert(`build-fix-escalated-${pr.agent || 'unassigned'}-${prDisplayId}`,
-            `# Build Fix Escalation\n\n` +
-            `**PR ${pr.id}**: ${pr.title || ''} on branch \`${pr.branch || 'unknown'}\` has failed **${pr.buildFixAttempts}** consecutive auto-fix attempts.\n` +
-            `**Last failure:** ${pr.buildFailReason || 'Check CI pipeline for details'}\n\n` +
-            `Auto-fix dispatch has been suspended. Please investigate manually.\n`
-          );
           try {
             const prPath = projectPrPath(project);
             mutatePullRequests(prPath, prs => {
@@ -2305,19 +2282,8 @@ async function discoverFromPrs(config, project) {
         } catch (e) { log('warn', 'increment build fix attempts: ' + e.message); }
       }
 
-      // Notify the author agent about the build failure
       if (pr.agent && !pr._buildFailNotified) {
-        let alertBody = `# Build Failure Notification\n\n` +
-          `**Your PR ${pr.id}**: ${pr.title || ''} on branch \`${pr.branch || 'unknown'}\` has a failing build.\n` +
-          `**Reason:** ${pr.buildFailReason || 'Check CI pipeline for details'}\n\n`;
-        if (pr.buildErrorLog) {
-          // Include first 30 lines of error log in notification (full log in fix agent prompt)
-          const logPreview = pr.buildErrorLog.split('\n').slice(0, 30).join('\n');
-          alertBody += `**Error preview:**\n\`\`\`\n${logPreview}\n\`\`\`\n\n`;
-        }
-        alertBody += `A fix agent has been dispatched to address this. Review the fix when complete.\n`;
-        writeInboxAlert(`build-fail-${pr.agent}-${prDisplayId}`, alertBody);
-        // Mark notified to prevent duplicate alerts
+        // Mark noted to prevent repeated processing; the fix dispatch and PR state carry the failure details.
         try {
           const prPath = projectPrPath(project);
           mutatePullRequests(prPath, prs => {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@yemi33/minions",
-  "version": "0.1.1610",
+  "version": "0.1.1612",
   "description": "Multi-agent AI dev team that runs from ~/.minions/ — five autonomous agents share a single engine, dashboard, and knowledge base",
   "bin": {
     "minions": "bin/minions.js"

package/playbooks/ask.md

@@ -28,7 +28,7 @@ The user has asked a question. Answer it thoroughly and clearly, writing your re
 - If the question is about "this" (ambiguous reference), check recent git history and agent activity for context
 
 ### 3. Write Your Answer
-Write your answer to `{{team_root}}/notes/inbox/{{agent_id}}-answer-{{task_id}}-{{date}}.md` with:
+After you have successfully answered the question, write your answer to `{{team_root}}/notes/inbox/{{agent_id}}-answer-{{task_id}}-{{date}}.md` with:
 
 - **Question**: (restate briefly)
 - **Answer**: (clear, direct answer)
@@ -36,11 +36,13 @@ Write your answer to `{{team_root}}/notes/inbox/{{agent_id}}-answer-{{task_id}}-
 
 Keep it concise but complete. Write for a senior engineer who wants the real answer, not fluff.
 
+If you cannot answer because the task is blocked, required context is missing, or research fails, do **not** write an inbox note. Report the blocker in your final response instead.
+
 ### 4. Status
 
 ## When to Stop
 
-Your task is complete once you have written your answer to the inbox file. Do NOT modify code, create PRs, or continue exploring beyond the question. Stop after writing.
+Your task is complete once you have written the successful answer to the inbox file. Do NOT modify code, create PRs, or continue exploring beyond the question. Stop after writing.
 
 ## Rules
 - Do NOT modify any code unless the question explicitly asks you to.

package/playbooks/build-and-test.md

@@ -73,7 +73,9 @@ If it is NOT a webapp (library, CLI tool, backend service without UI), skip this
 
 ## Output Format
 
-Write your findings to: `{{team_root}}/notes/inbox/{{agent_id}}-bt-{{pr_number}}-{{date}}.md`
+Write your findings to `{{team_root}}/notes/inbox/{{agent_id}}-bt-{{pr_number}}-{{date}}.md` **only after a successful verification run**: the build passed, required tests passed, and any applicable local server is running or not applicable.
+
+If the build fails, tests fail, dependency setup fails, or a required local server cannot start, do **not** write an inbox note. Follow the failure handling below and report the failure in your final response instead.
 
 Structure your report exactly like this:
 
@@ -85,16 +87,16 @@ Structure your report exactly like this:
 **Project:** {{project_name}}
 
 ### Build
-- Status: PASS / FAIL
+- Status: PASS
 - Notes: (any warnings or issues)
 
 ### Tests
-- Status: PASS / FAIL / SKIPPED
-- Results: X passed, Y failed, Z skipped
-- Failed tests: (list if any)
+- Status: PASS / SKIPPED
+- Results: X passed, 0 failed, Z skipped
+- Failed tests: none
 
 ### Local Server
-- Status: RUNNING / NOT_APPLICABLE / FAILED
+- Status: RUNNING / NOT_APPLICABLE
 - URL: http://localhost:XXXX (if running)
 - Run Command: `cd <absolute-path> && <command>`
 
@@ -152,4 +154,4 @@ Leave the worktree in place at `{{project_path}}/../worktrees/bt-{{pr_number}}`
 
 ## When to Stop
 
-Your task is complete once you have: (1) built the project, (2) run tests, (3) started the app if applicable, and (4) written your findings to the inbox file. Stop after writing findings.
+Your task is complete once you have: (1) built the project, (2) run tests, (3) started the app if applicable, and (4) written the success findings to the inbox file. If verification failed, stop after filing the failure work item when applicable and reporting the failure in your final response; do not write an inbox file.

package/playbooks/explore.md

@@ -31,7 +31,7 @@ Explore the codebase area specified in the task description. Your primary goal i
 - What conventions are followed?
 
 ### 4. Document Findings
-Write your findings to `{{team_root}}/notes/inbox/{{agent_id}}-explore-{{task_id}}-{{date}}.md` with these sections:
+After the requested exploration succeeds, write your findings to `{{team_root}}/notes/inbox/{{agent_id}}-explore-{{task_id}}-{{date}}.md` with these sections:
 - **Area Explored**: what you looked at
 - **Architecture**: how it works
 - **Patterns**: conventions and patterns found
@@ -40,6 +40,8 @@ Write your findings to `{{team_root}}/notes/inbox/{{agent_id}}-explore-{{task_id
 - **Recommendations**: suggestions for the team
 - **Source References**: for EVERY finding, include the source — file paths, line numbers, PR URLs, API endpoints, config keys. Format: `(source: path/to/file.ts:42)` or `(source: PR-12345)`. This is critical — other agents and humans need to verify your findings.
 
+If exploration is blocked or fails before you can produce sourced findings, do **not** write an inbox note. Report the blocker in your final response instead.
+
 ### 5. Create Deliverable (if the task asks for one)
 If the task asks you to write a design doc, architecture doc, or any durable artifact:
 1. Write the document in the current working directory (e.g., `docs/design-<topic>.md`)
@@ -64,7 +66,7 @@ Use subagents only for genuinely parallel, independent tasks. For reading files,
 
 ## When to Stop
 
-Your task is complete once you have written your findings to the inbox file. Do NOT continue reading additional files, exploring tangential areas, or producing extra analysis beyond what was asked. Write your findings and stop.
+Your task is complete once you have written the successful findings to the inbox file. Do NOT continue reading additional files, exploring tangential areas, or producing extra analysis beyond what was asked. Write your findings and stop.
 
 ## Team Decisions
 {{notes_content}}

package/playbooks/shared-rules.md

@@ -24,6 +24,7 @@ Your context window may be compacted or summarized mid-task by Claude's automati
 - Do NOT write to `agents/*/status.json` — the engine manages your status automatically.
 - Do NOT remove worktrees — the engine handles cleanup automatically.
 - Do NOT checkout branches in the main working tree — use worktrees or `git diff`/`git show`.
+- Treat `notes/inbox/` writes as success artifacts only. If the task fails, is blocked, is cancelled, or ends partial, do **not** create an inbox note; report the failure in your final response, completion block, PR/work-item comment, or other task-specific failure channel instead.
 - Read `notes.md` for team rules and decisions before starting.
 - **Check team memory first, then look outside.** Before researching from scratch, check what the team already knows — in this order:
   1. `pinned.md` — critical context flagged by the human teammate (READ FIRST)

package/playbooks/test.md

@@ -53,12 +53,14 @@ Example: `cd C:/Users/you/my-project && yarn dev`
 
 The agent process terminates after completion, so any dev server you start will die with it. The user needs this command to run it themselves.
 
-## After Completion
+## After Successful Completion
 
-Write your findings to: `{{team_root}}/notes/inbox/{{agent_id}}-{{item_id}}-{{date}}.md`
+Write your findings to `{{team_root}}/notes/inbox/{{agent_id}}-{{item_id}}-{{date}}.md` **only after a successful run**: all requested build/test/run steps completed, required checks passed, and any required PR was created.
+
+If any requested build, test, run, or file-change step fails or is blocked, do **not** write an inbox note. Report the failure clearly in your final response, including the command, error summary, and any next action needed.
 
 Include:
-- Build status (pass/fail)
+- Build status
 - Test results if applicable
 - Any errors or warnings
 - The run command (absolute paths, copy-pasteable from any terminal)
@@ -67,6 +69,6 @@ Include:
 
 ## When to Stop
 
-Your task is complete once you have run the tests, written findings to the inbox file, and (if the task involved file changes) created and submitted a PR. Stop after writing findings.
+Your task is complete once you have run the tests, written the success findings to the inbox file, and (if the task involved file changes) created and submitted a PR. If the run failed or was blocked, stop after reporting the failure in your final response; do not write an inbox file.
 
 Do NOT remove worktrees — the engine handles cleanup automatically.

package/playbooks/verify.md

@@ -77,9 +77,13 @@ If the project has no runnable application, skip this step and note that in the
 
 ## Step 5: Write the Verification Report and Testing Guide
 
-Create the guide in TWO locations:
-1. **Permanent location** (linked from dashboard): `{{team_root}}/prd/guides/verify-{{plan_slug}}.md`
-2. **Inbox copy** (for team consolidation): `{{team_root}}/notes/inbox/verify-{{plan_slug}}.md`
+Always create the permanent guide linked from the dashboard:
+1. **Permanent location**: `{{team_root}}/prd/guides/verify-{{plan_slug}}.md`
+
+Create the inbox copy only after a successful verification run:
+2. **Inbox copy**: `{{team_root}}/notes/inbox/verify-{{plan_slug}}.md`
+
+A successful verification run means every required project build/test passed or was legitimately not applicable, each runnable app was started and smoke-checked when required, and the required E2E PRs were created or updated. If verification is failed, blocked, or partial, do **not** create the inbox copy; record the details in the permanent guide and final response instead.
 
 **Be transparent.** The guide must clearly state what was built, what was tested, what passed, what failed, and what still needs human verification.
 
@@ -162,7 +166,7 @@ Use subagents only for genuinely parallel, independent build/test tasks on separ
 
 ## When to Stop
 
-Your task is complete once you have: (1) merged dependency branches, (2) built and tested, (3) written the verification report to both locations, and (4) created the E2E PR(s).
+Your task is complete once you have: (1) merged dependency branches, (2) built and tested, (3) written the verification report to the permanent guide, (4) written the inbox copy only if verification succeeded, and (5) created the E2E PR(s).
 
 **IMPORTANT: Your final message MUST include the E2E PR URL(s) so the engine can track them.** Example final message:
 

package/playbooks/work-item.md

@@ -45,9 +45,11 @@ Keep branch names lowercase, use hyphens, max 60 chars.
 
 Do NOT remove the worktree — the engine handles cleanup automatically.
 
-## After Completion
+## After Successful Completion
 
-Write your findings to: `{{team_root}}/notes/inbox/{{agent_id}}-{{item_id}}-{{date}}.md`
+Write your findings to `{{team_root}}/notes/inbox/{{agent_id}}-{{item_id}}-{{date}}.md` only after the work item succeeds: build/tests pass, the branch is pushed, and the PR is created.
+
+If you stop because the task failed, is blocked, or is only partially complete, do **not** write an inbox note. Put the failure details in your final response and in any required PR/work-item comment instead.
 
 ## Handling Merge Conflicts
 If you encounter merge conflicts during push or PR creation: