宜博 / lemonboss

1. Rewrote master/agent.md: strict anti-overreach rules, cleaned git merge conflicts
2. Enhanced agent-creator/agent.md: force Write tool for new files, mandatory task report
3. Implemented task/report callback: auto-wakes sender agent to continue next steps
   - Added .tasks/ metadata persistence for dispatch→report traceability
   - task/report sends [TASK_CALLBACK] to dispatching agent on done/failed

- company.js: call heartbeatService._setupAgentHeartbeat after hire
- CEO agent.md: enforce 3-step continuous execution in same turn
- All CXO heartbeat intervals: 5 minutes
- task dispatch message: self-drive directive, use lemon CLI
- bin/lemon.js: add 'task report' command

…ter Agent cognitive auto-recovery loop

- Refactored CompanyDashboard into read-only Observer UI
- Enforced strict companyId:roleId namespacing across Node APIs and AgentManager
- Added dynamic <env_status> injection to PromptBuilder for robust context
- Introduced bin/lemon.js CLI for structured LLM tool execution
- Implemented in-stream error interceptors for implicit silent retry loop

…nti-confirmation prompts, cron dedup fix

- CompanyDashboard.vue: full company management UI
  - Company list grid with status badges
  - Company detail view with tabs (Agents/Pipelines/Goals)
  - Create company modal
  - Hire agent modal with template/training feedback
  - Template library slide-out panel
  - Pipeline flow visualization (step nodes with arrows)
  - Pipeline run history with status
  - OKR goals list
- Route: /company registered in Vue Router
- Build verified: all assets compiled successfully

- CompanyHeartbeatService: company-level heartbeat with timer registration
- Auto-trigger: cron-based pipeline scheduling (intervalHours)
- Active run driving: dispatch tasks to agents, timeout detection
- CEO heartbeat: periodic company status check
- Pipeline cron field support in definitions
- Registered in server startup alongside agent heartbeat
- E2E test: 113/113 passed (all phases + autonomous loop)

- PipelineService: define pipelines with ordered agent-assigned steps
- Run lifecycle: start → markStepStarted → advanceStep → complete
- Step output automatically flows as next step input
- Pipeline CRUD + Run management REST API routes
- E2E test: 106/106 passed (Phase 1 + 2 + 3 + 4)

- GoalsService refactored: constructor accepts optional companyId
  - With companyId: goals stored in data/{companyId}/goals/
  - Without: backward-compatible data/goals/ (global)
- Company Goals REST API: full OKR CRUD under /api/company/:id/goals
- OKR isolation verified: different companies have independent goal spaces
- E2E test: 81/81 passed (Phase 1 + 2 + 3)

- scaffoldNewAgent: create agent skeleton via AgentScaffold directly
- depositTemplate: benchmark gating (≥95) + auto-deposit to templates/
- hireWithTraining: full flow (template lookup → scaffold fallback)
- Company route: /hire uses hireWithTraining, new /deposit-template endpoint
- E2E test: 69/69 passed (Phase 1 + Phase 2)

- CompanyService: CRUD (create/get/list/update/shutdown)
- Agent Hiring Pipeline: template lookup → copy → context injection
- Template registry (_registry.json) with CEO and qa-judge initial templates
- resolveAgentDir(companyId, agentId) unified agent directory resolution
- Company REST API routes (/api/company/*)
- E2E test: 46/46 passed

- Extracted HeartbeatService to periodically wake up agents flagged with 'heartbeat.enabled'
- Sends a silent [SYSTEM HEARTBEAT] message to trigger self-checks (e.g. OKR state validation)
- Automatically integrated service dispatch hook in server startup sequence
- Configured qa-judge to run autonomous heartbeat every 10 min

…ze Chinese keyword extraction

- Fix: The long trailing history in todo.md drowned out the actual current task keywords because English tokens (e.g. TASK, P2, from) padded the 30-keyword limit beforehand.
- Updated extractKeywords: Process Chinese characters first as they carry more weight for NLP context intent.
- Added userMessage prefix to searchText in buildSystemPrompt so newly dispatched tasks dominate keyword extraction.
- End-to-end e2e test created and passed (27/27 asserts passed)

…sistant with troubleshooting capabilities

Phase 3 features:
- autoReflect: auto-trigger on complex conversations (toolCount>=3 or duration>=60s)
  - Extracts patterns from successful conversations
  - Extracts lessons from failed conversations
  - Auto-solidifies cerebellum plans from 5+ tool-call successes
  - Writes reflection logs to memory/reflections/
  - Updates memory.md 'verified patterns' and 'lessons learned' sections
- consolidateMemory: merge short-term sessions older than 7 days into long-term/
  - Extracts frequent keywords across merged sessions
  - Writes consolidated summary to long-term/consolidate-{date}.md
  - Updates long-term/patterns.md with high-frequency patterns
- generateDailyReport: aggregate today's sessions and reflections into daily/{date}.md
- API endpoints: POST /:agentId/memory/consolidate, POST /:agentId/memory/daily-report

Test results: 16/16 passed (Phase 3 test suite)

- Hippocampus keyword matching engine (extractKeywords + score)
- Cerebellum plan storage (saveCerebellumPlan)
- Auto-sync hippocampus-index.json + hippocampus.md
- POST /:agentId/memory/cerebellum API endpoint
- buildSystemPrompt injects matched plans from hippocampus
- Use count & lastUsed tracking on plan access
- Chinese n-gram + English word tokenizer with stop words
- Fix: score formula uses entry keyword count as denominator
- Fix: useCount updates original index entry not spread copy
- Verified: OKR plan matched + injected into CEO prompt

- New: server/services/memory-service.js
  - saveSessionSummary(): auto-save session summary to short-term/
  - loadRecentMemory(): inject last 5 sessions into system prompt
  - appendToMemoryMd(): auto-append key facts with size control
- Modified: server/routes/agent-chat.js
  - buildSystemPrompt() now loads recent short-term sessions
  - Conversation completion triggers auto memory save
- Verified: CEO Agent conversation creates session file + updates memory.md

- Fix loadIdentity regex to handle **bold** markdown formatting
- Add fallback to config.json name when identity.md has no 名称 field
- Add 名称 field to Digester/Evolver/Innovator identity.md

- New: docs/system-design/21-fde-pipeline-architecture.md
  Complete FDE training pipeline documentation with:
  - Triple-layer architecture (Hunt→Digest→Evolve)
  - Dual-path output convention (workspace + agents)
  - Dispatch protocol examples
  - buildSystemPrompt skills scanning fix details
  - CEO training verification results (92/100)
- Update: architecture-v5.md
  - Phase 3 OKR checkboxes updated (Goals API done)
  - Phase 3.5 FDE added as completed milestone
  - 3 new architecture decisions (#10-12)
  - File path index expanded with meta agents

Hunt→Digest→Evolving full pipeline verified:
- Hunter: 1461 lines (knowhow + toplist + 6 crawl samples)
- Digester: benchmark_v1 + 5 good/1 bad testdata + feature analysis
- Evolver: 3 iterations (v1:40 → v2:87 → v3:92), CEO SKILL.md v2.3.0
- Dual-path sync: workspace/ + agents/ both populated
- All task reports successfully sent to Innovator

- Hunter/Digester/Evolver now write to both locations:
  - workspace/{targetAgentId}/hunt|digest|evolve/ (shared)
  - agents/{targetAgentId}/hunt|digest|evolve/ (agent asset)
- Fix buildSystemPrompt to scan ALL .claude/skills/ subdirs (critical)
- Verified Hunter test: 22k chars prompt, 1461 lines output