land-pr

Pass

Audited by Gen Agent Trust Hub on Nov 3, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from pull request bodies, review comments, and CI logs to make merge decisions.
  • Ingestion points: Fetches pull request data, review comments, and CI logs using gh api and gh run view as described in SKILL.md (Steps 1, 2, and 3).
  • Boundary markers: The skill includes explicit instructions to only merge upon receiving a specific "owner's word" (e.g., "merge", "ship it") and to check for AGENTS.md prohibitions.
  • Capability inventory: The skill can merge pull requests (gh pr merge) and resolve review threads via GraphQL mutations.
  • Sanitization: A "gate" mechanism is described in Step 6 to ensure the merge command is executed in isolation with strict shell options to prevent command injection.
  • [COMMAND_EXECUTION]: The skill uses various command-line tools to interact with the repository and the GitHub API.
  • Evidence: Frequent use of gh api, gh api graphql, gh pr merge, and git log throughout the operational steps in SKILL.md.
  • [DATA_EXPOSURE]: The skill accesses pull request metadata and build logs, which is a requirement for its primary functionality.
  • Evidence: Accesses failing CI logs and PR comments to facilitate automated fixes and feedback loops.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 3, 2026, 04:48 PM