Audit-grade memory backbone for agent teams. Bi-temporal facts (event time + transaction time, with recall(as_of=...) replay), 6-step deterministic retrieval (no LLM in the critical path), conversation ingest with speaker-locked dual-pass extraction, per-tenant Postgres row-level security, and Ed25519-signed provenance. Postgres + pgvector + Neo4j defaults.
Scores are estimated from documentation until an agent is runtime-measured. Methodology v2.6.0 rev 1.27.
Week of 2026-10-12 · Automated Tier 1
Methodology v2.6.0Independent benchmark across four dimensions.
2 improvement opportunities are below their benchmark — these are estimates from documentation; sign in to see how to raise your estimated score.Estimated from documentation
See your ranked fixes →Automated Tier 1 · Week of 2026-10-12
Assessor agreement: High — independent review stages closely agreed
Score reflects an independent capability assessment. Community signals (stars, contributors) appear separately below as adoption indicators that complement — but do not replace — the score.
Scan scope: the agent's public GitHub repository. The deployed service may differ from the scanned source.
No scan history available yet.
No current findings.
No current findings.
No current findings.
Independent adoption indicators from GitHub. These complement — but do not replace — the capability score above.