archive-agent-notes
Pass
Audited by Gen Agent Trust Hub on Aug 20, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes external markdown files (Agent Notes) to perform semantic classification, creating a surface for indirect prompt injection if those files contain malicious instructions.
- Ingestion points: The agent reads
foo.mdandfoo.zh.mdfrom the repository to evaluate rationales and implement archive status changes. - Boundary markers: There are no boundary markers or instructions telling the agent to treat the content of the notes as data rather than instructions.
- Capability inventory: The agent has capabilities to move files, modify existing documentation, and execute repository scripts.
- Sanitization: No sanitization or filtering of the note content is performed prior to the agent's evaluation.
- [COMMAND_EXECUTION]: The skill instructs the agent to run local repository scripts and tools for validation purposes.
- Evidence: The instructions explicitly direct the agent to "Run the repository's archive seal gate, then its documentation gates and git diff --check".
Audit Metadata