local-conversation-history
Pass
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill directs the agent to ingest data from local conversation stores such as ~/.claude/history-sources.json. Ingestion points: Conversation history directories and registry files. Boundary markers: Instructions mention preserving session boundaries and excluding current session data to prevent self-referential confusion. Capability inventory: The skill routes to other skills with file-read and history-parsing capabilities. Sanitization: Not specified at this routing layer.
- [SAFE]: Interaction with local file paths like ~/.claude is limited to the skill's stated purpose of managing AI history and does not involve unauthorized data access or exfiltration.
Audit Metadata