72c548b729
test: expand coverage from 20 to 92 tests, migrate to uv
...
- Migrate from setuptools to hatchling build backend
- Add dependency-groups (PEP 735) for dev tooling (pytest, ruff)
- Remove redundant requirements.txt in favor of uv.lock
- Fix __version__ mismatch (2.0.0 -> 3.0.0 to match pyproject.toml)
New test files:
- conftest.py: shared fixtures (isolated palace, KG, ChromaDB collection)
- test_knowledge_graph.py: 17 tests (entity CRUD, temporal queries, timeline)
- test_mcp_server.py: 25 tests (protocol dispatch, read/write/KG/diary tools)
- test_searcher.py: 7 tests (search_memories API, filters, error handling)
- test_dialect.py: 13 tests (AAAK compression, entity/emotion detection, zettel encoding)
All 92 tests pass on Python 3.13 with chromadb 0.6.3.
2026-04-07 17:55:10 -03:00
186bb2e3d1
fix: shell injection in hooks, Claude Code mining, chromadb pin
...
- hooks/mempal_save_hook.sh: pass $TRANSCRIPT_PATH as sys.argv
instead of interpolating into python -c string (fixes #110 )
- normalize.py: accept type "user" in addition to "human" for
Claude Code JSONL sessions (fixes #111 )
- convo_miner.py: skip tool-results/, memory/ dirs and .meta.json
files when scanning for conversations (fixes #111 )
- pyproject.toml: pin chromadb>=0.4.0,<1 to avoid crashing 1.x
builds on macOS ARM64 (fixes #100 )
2026-04-07 11:45:51 -07:00
068dbd9a7b
MemPalace: palace architecture, AAAK compression, knowledge graph
...
The memory system:
- Palace structure: Wings (people/projects) → Rooms (topics) → Closets (AAAK compressed) → Drawers (verbatim transcripts)
- Halls connect related rooms within a wing
- Tunnels cross-reference rooms across wings
- AAAK: 30x lossless compression dialect for AI agents
- Knowledge graph: temporal entity-relationship triples (SQLite)
- Palace graph: room-based navigation with tunnel detection
- MCP server: 19 tools — search, graph traversal, agent diary, AAAK auto-teach
- Onboarding: guided setup generates wing config + AAAK entity registry
- Contradiction detection: catches wrong pronouns, names, ages
- Auto-save hooks for Claude Code
96.6% Recall@5 on LongMemEval — highest zero-API score published.
100% with optional Haiku rerank (500/500).
Local. Free. No API key required.
2026-04-04 18:16:04 -07:00
Igor Lins e Silva