update-agents-md
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands
wc -lto verify file length constraints onAGENTS.mdandREADME.md, andnpx markdownlint-clito perform linting. These are standard development tools used for validation. - [INDIRECT_PROMPT_INJECTION]: The skill creates an attack surface for indirect prompt injection by instructing the agent to read and modify documentation files that could contain untrusted instructions.
- Ingestion points: The agent reads the content of
AGENTS.mdandREADME.mdto perform restructuring and maintenance tasks. - Boundary markers: The instructions do not define any delimiters or explicit safety warnings to ignore instructions embedded within the target files.
- Capability inventory: The agent is authorized to execute shell commands (
wc,npx) and write changes back to the file system. - Sanitization: There is no evidence of sanitization, filtering, or validation of the input content before the agent processes it.
Audit Metadata