Structural memory for AI encoding agents
syncable-dev/memtrace-public · skills
Structural memory for AI coding agents.
★ 468GitHub stars
41Forks
2026-08-17Last updated
PythonLanguage
NOASSERTIONLicense
Key features
- Each agent reads the same call graph, sees the same blast radius, inherits the same temporal history.
- Memtrace is a bi-temporal episodic structural knowledge graph.
- Graphiti processes data through add_episode(), which triggers multiple LLM calls per episode — entity extraction, relationship resolution, deduplication.
- find_code — hybrid BM25 + semantic + RRF
- Replay any refactor with full causal awareness.
Requirements
- Skills are workflow prompts that teach the agent how to chain tools.
- Kiro does not use SKILL.md, so Memtrace writes equivalent auto steering files instead.
Install command
dsh plugin --profile web add github:syncable-dev/memtrace-public