test-first-by-evidence

Pass

Audited by Gen Agent Trust Hub on Sep 11, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill directs the agent to identify and execute test commands from the user's repository, which creates a surface for potential script execution if the repository is untrusted.\n
  • Ingestion points: The agent is instructed to read the project's configuration (e.g., package.json) and existing tests to determine the correct command for running tests (referenced in SKILL.md and rules/watch-it-fail.md).\n
  • Boundary markers: There are no instructions or boundary markers provided to verify the safety or integrity of the repository's scripts before execution.\n
  • Capability inventory: The skill relies on the agent's ability to execute shell commands to facilitate the TDD cycle.\n
  • Sanitization: The instructions do not require the agent to sanitize or validate the command strings retrieved from the project environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 11, 2026, 01:26 PM
Security Audit — agent-trust-hub — test-first-by-evidence