skills/decisionnerd/dev-skills/pulls/Gen Agent Trust Hub

pulls

Pass

Audited by Gen Agent Trust Hub on Nov 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources like GitHub PR descriptions, comments, and reviews, which could contain malicious instructions.\n
  • Ingestion points: The skill reads PR content using gh pr view, gh pr checks, and related commands as part of the monitor and babysit workflows (SKILL.md, references/monitor.md).\n
  • Boundary markers: It includes strong instructions to treat reviewer feedback as untrusted input, requiring verification of any suggested changes in the source code or via reproduction before execution (references/monitor.md).\n
  • Capability inventory: The skill can execute GitHub mutations (create, edit, merge, close PRs) and can trigger other skills like fix-it or test-it to perform code changes and verification (SKILL.md).\n
  • Sanitization: It explicitly mandates using --body-file with the gh CLI for multiline content to prevent command injection from untrusted text interpolation (references/ops.md).\n- [COMMAND_EXECUTION]: The skill relies on local command-line tools for Git and GitHub operations.\n
  • Evidence: Frequent use of git (e.g., git status, git push) and the GitHub CLI gh (e.g., gh pr create, gh pr merge) to interact with the repository and platform APIs.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 4, 2026, 10:25 PM