 Profa
|
b70ef3566e
|
fix: v1.8.4 - Learning persistence and hook schema fixes
This release fixes critical issues preventing learning data persistence
and pre-edit hook failures in QE agents running as Claude Code subagents.
## Fixed Issues
### 1. Memory Manager Fragmentation (#66)
- Multiple isolated SwarmMemoryManager instances caused data fragmentation
- Implemented singleton pattern via MemoryManagerFactory
- All components now share same database connection
- Data properly persists across agent executions
### 2. Database Closure on Exit (#66)
- sql.js (WASM SQLite) requires explicit close() to persist
- Added process exit handlers (SIGINT, SIGTERM, exit, uncaughtException)
- Ensures data survives process termination
### 3. Memory Backup Schema Mismatch
- memory-backup.ts referenced 'namespace' but schema uses 'partition'
- Also referenced 'timestamp' instead of 'createdAt'
- Fixed all 5 affected lines (80, 84, 85, 132, 134)
- Pre-edit hooks now work without "no such column: namespace" errors
## Changes
### Added
- src/core/memory/MemoryManagerFactory.ts - Singleton pattern implementation
- Process exit handlers for graceful database closure
- getSharedMemoryManager(), initializeSharedMemoryManager(), resolveDbPath()
### Modified
- src/mcp/server.ts - Use singleton memory manager
- src/mcp/services/AgentRegistry.ts - Use singleton memory manager
- src/mcp/handlers/phase2/Phase2Tools.ts - Use singleton memory manager
- src/mcp/handlers/memory/memory-backup.ts - Fixed schema column names
- src/core/memory/index.ts - Export factory functions
### Documentation
- Fixed GitHub URLs from ruvnet/agentic-qe-cf to proffesor-for-testing/agentic-qe
- Updated CLAUDE.md, skills, and guides
## Testing
- Verified singleton pattern with test project
- Confirmed learning data persists (3 experiences, 3 Q-values, 3 patterns)
- No temp_*.db files created
- Database grows correctly (26MB with new data)
- Average reward: 0.93 (excellent performance)
Closes #66
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
|
2025-11-20 08:52:28 +00:00 |
|
|
Profa
|
b3d2018ee2
|
feat: v1.8.0 - Quality Hardening & Pattern Isolation
Critical bug fixes and code quality improvements from brutal-honest-review.
## Critical Fixes (9/10 completed, 1 deferred)
1. Fixed Agent Pattern Isolation (CRITICAL)
- Added SwarmMemoryManager.queryPatternsByAgent()
- SQL filtering by agent_id in metadata
- Each agent now isolated
2. Fixed Async Method Cascade
- Changed getPatterns() to async
- Fixed 10 callers across 6 files
- Build passes
3. Fixed Misleading Logging
- Added isRealAgentDB() detection
- Honest mock vs real reporting
4. Consolidated Embedding Generation
- Created EmbeddingGenerator utility
- Removed 50+ duplicate lines
5. Added AgentDB Initialization Checks
- Empty DB detection
- HNSW index verification
6. Integration Test Suite
- 26 new test cases
- Error path coverage
7. Repository Cleanup
- Removed 19 temp test files
8. New QE Skill: sherlock-review
- Evidence-based investigative review
- 38 QE skills total
9. Deferred: Wire Up Real Test Execution
- Deferred to v1.9.0 (documented)
## Statistics
- Fixes: 9/10 (90%, 1 deferred)
- Files Modified: 16
- Files Created: 3
- Files Deleted: 19
- Integration Tests: 26
- Build: PASSING
## Breaking Changes
- getPatterns() now async - add await
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
|
2025-11-17 12:32:03 +00:00 |
|