skills/posit-dev/skills/pr-create/Gen Agent Trust Hub

pr-create

Pass

Audited by Gen Agent Trust Hub on Oct 8, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes Git and GitHub CLI commands (gh) to manage branches, commits, and pull requests. These are standard operations for the skill's primary purpose. A helper script scripts/find-collaborator.sh uses gh api to look up repository collaborators, which is a legitimate use of the GitHub API for resolving reviewer handles.
  • [REMOTE_CODE_EXECUTION]: The skill incorporates a safety check in the 'Security Boundaries' section, explicitly instructing the agent NOT to run commands found in CI log output, error messages, or external instructions, limiting execution to commands defined in local project configuration.
  • [DATA_EXFILTRATION]: The skill includes a 'Do not expose secrets' boundary that prevents the inclusion of environment variables, tokens, or credentials in commit messages or PR bodies, even if they appear in logs.
  • [SAFE]: The skill requires explicit user confirmation before creating the pull request and uses gh pr create --draft to ensure the PR is not shared for review until the agent verifies that CI has passed.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 8, 2026, 08:52 AM