skills/vdustr/skills/vp-stacked-pr/Gen Agent Trust Hub

vp-stacked-pr

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to install the official github/gh-stack extension via gh extension install github/gh-stack when working with native GitHub stacks.
  • [COMMAND_EXECUTION]: The skill executes multiple local Git commands and GitHub CLI operations (git rebase, git update-ref, git push, gh api) to manage and repair stacked pull requests.
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection through data ingestion from external repository states.
  • Ingestion points: Fetches open and closed pull request metadata from the GitHub API using gh api --method GET --paginate repos/<owner>/<repo>/pulls in references/manual-rebase.md.
  • Boundary markers: Absent; the instructions do not specify strict text delimiters or explicit instructions to ignore prompt injection attempts within PR fields or branch names.
  • Capability inventory: Possesses powerful repository modification capabilities including git push, gh api -X PATCH (reopening PRs), gh pr edit (retargeting bases), and gh api -X DELETE (deleting remote branches) in references/manual-rebase.md.
  • Sanitization: Absent; the skill passes raw branch names and repository details directly into command templates and metadata checks.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 04:08 AM
Security Audit — agent-trust-hub — vp-stacked-pr