memory
Pass
Audited by Gen Agent Trust Hub on Apr 13, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill manages a persistent memory system that stores user facts and conversation history, which creates a surface for indirect prompt injection. 1. Ingestion points: The agent reads from SUMMARY.md, PROFILE.md, and SOUL.md as described in SKILL.md. 2. Boundary markers: The skill does not specify any delimiters or instructions to treat memory content as untrusted. 3. Capability inventory: The agent uses edit_file and bash (grep) to interact with the file system. 4. Sanitization: No sanitization or validation of the stored memory content is performed.
- [COMMAND_EXECUTION]: The skill instructs the agent to use the grep command via a shell to search interaction history. Evidence: grep -i "keyword" memory/SUMMARY.md.
Audit Metadata