grading-claude-agents-md

Pass

Audited by Gen Agent Trust Hub on Mar 25, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes CLAUDE.md and AGENTS.md files from the user's project, which creates a surface for indirect prompt injection where malicious instructions embedded in these files could override agent behavior during analysis.
  • Ingestion points: Project configuration files (CLAUDE.md, AGENTS.md).
  • Boundary markers: No delimiters or ignore instructions are used when reading the files.
  • Capability inventory: The skill can read/write files and execute shell commands.
  • Sanitization: No sanitization is applied to the input content.
  • Mitigation: The skill enforces a human-in-the-loop approval process before any changes are committed to the filesystem.
  • [COMMAND_EXECUTION]: The skill instructs the agent to perform shell operations such as file size measurement (wc), directory listings (ls), and project-specific tasks (npm test, pnpm build) as part of its grading and validation workflow.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 25, 2026, 12:17 PM