Skills
Navigate
skills/heyeddi-com/heyeddi-skills/heyeddi-pr-respond/Gen Agent Trust Hub

heyeddi-pr-respond

Pass

Audited by Gen Agent Trust Hub on Sep 11, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes PR comments which are untrusted external data.
  • Ingestion points: scripts/fetch_pr_comments.py (fetches comment and review data from GitHub via gh CLI).
  • Boundary markers: scripts/_untrusted_doc.py wraps untrusted content with <<<UNTRUSTED_EXTERNAL_CONTENT>>> markers and explicit instructions for the agent to treat it as DATA only.
  • Capability inventory: The skill can execute git and gh commands, as well as project-specific test runners (npm, pytest, make, go, cargo) via scripts/discover_and_verify.py.
  • Sanitization: scripts/_untrusted_doc.py includes wrap_comment_bodies to recursively wrap free-text fields like body and diff_hunk.
  • [COMMAND_EXECUTION]: The skill executes build and test commands discovered in the project repository.
  • The scripts/discover_and_verify.py tool identifies and runs commands such as npm run test, pytest, or make test based on files like package.json or Makefile.
  • Execution is performed via scripts/_skill_cli.py using subprocess.run with shell=False and absolute executable paths, which is a secure practice.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 11, 2026, 03:38 AM
Security Audit — agent-trust-hub — heyeddi-pr-respond