implement

Pass

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data from project documentation, specifications, and tickets to guide code implementation and verification.
  • Ingestion points: Reads from files in the .agents/projects/<project>/ directory, including CONTEXT.md, LINKS.md, todo.md, approved specs, and ticket drafts.
  • Boundary markers: No explicit delimiters or instructions are provided to the agent to treat external content as data rather than instructions.
  • Capability inventory: The skill allows for source code modification, execution of local test suites and coverage tools, and posting comments to external platforms like Jira or GitHub.
  • Sanitization: The instructions do not include steps for sanitizing or validating the content of the ingested project artifacts before they influence the implementation process.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 27, 2026, 06:37 AM
Security Audit — agent-trust-hub — implement