pr-creator

Pass

Audited by Gen Agent Trust Hub on Aug 10, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements a standard software development workflow using well-known tools (git, npm, gh).- [PROMPT_INJECTION]: The skill involves an indirect prompt injection surface by reading external pull request templates (.github/pull_request_template.md).
  • Ingestion points: Reads template files from the .github/ directory in the local workspace.
  • Boundary markers: None explicitly defined for the template content itself.
  • Capability inventory: The skill has access to git (branching, status, pushing), npm (preflight checks), gh (PR creation), and rm (file cleanup).
  • Sanitization: The skill follows security best practices by writing the PR body to a temporary file and using the --body-file flag with the gh CLI, which prevents potential shell injection or escaping issues that could arise from malicious or malformed text in a template.
  • [COMMAND_EXECUTION]: All shell commands (git checkout, git add, git commit, npm run preflight, git push, gh pr create) are restricted to standard development tasks and include explicit checks to prevent execution on the protected main branch.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 10, 2026, 05:31 AM
Security Audit — agent-trust-hub — pr-creator