Agent Memory
PassAudited by ClawScan on May 10, 2026.
Overview
Agent Memory is a disclosed local memory database for agents; no exfiltration or credential use is shown, but saved memories can persist and influence later sessions.
This looks purpose-aligned for local persistent memory. Before installing, decide what kinds of facts the agent is allowed to save, avoid secrets or highly sensitive personal data, and periodically inspect or clear the local memory database.
Findings (1)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Private details, outdated facts, or mistaken memories could be saved and later influence the agent's behavior.
This shows the skill intentionally stores conversation-derived facts in a persistent local database and reuses them across future agent sessions.
Add to your AGENTS.md or HEARTBEAT.md... On session start... Recall relevant facts... On session end... Extract durable facts from conversation... Default: `~/.agent-memory/memory.db`
Store only durable non-secret information, periodically review or delete memories, and consider using separate database paths for different projects or users.