append-text-to-file
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill retrieves and displays existing file content to establish context before performing an append operation. This creates a surface for indirect prompt injection if the file being read contains malicious instructions that the agent might follow.\n
- Ingestion points: Step 2 of the Procedure in SKILL.md calls for retrieving current file content.\n
- Boundary markers: Absent. There are no instructions to treat the file content as untrusted data or wrap it in delimiters.\n
- Capability inventory: The skill uses the write_file tool to modify the file system.\n
- Sanitization: No evidence of sanitization or validation of the retrieved file content.\n- [COMMAND_EXECUTION]: The Reference Commands section contains a shell-based command:
cat {{FILE_PATH}} && echo '{{NEW_CONTENT}}' >> {{FILE_PATH}}. If the agent executes this literally instead of using the recommended internal tools, it could be vulnerable to command injection if user-provided paths or content contain shell metacharacters. The skill notes correctly recommend using the write_file tool to mitigate this risk.
Audit Metadata