prove-a-pr

Pass

Audited by Gen Agent Trust Hub on Aug 26, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes shell commands to manage project builds and run test suites. Evidence includes commands such as pnpm --filter @openwork/types build and pnpm evals:e2e --publish used to prepare environments and report results.
  • [COMMAND_EXECUTION]: The agent uses command-line tools to interact with Git and GitHub to verify pull request state. Evidence includes gh pr view and git log used to inspect branches and commit history.
  • [PROMPT_INJECTION]: The skill processes data from GitHub pull requests, which is an external and potentially untrusted source (Indirect Prompt Injection surface). 1. Ingestion points: The skill uses gh pr view to retrieve metadata about pull requests (SKILL.md). 2. Boundary markers: None identified in the prompt instructions for handling PR metadata. 3. Capability inventory: The skill has shell execution capabilities via pnpm, git, and gh (SKILL.md). 4. Sanitization: No explicit sanitization or validation of the PR metadata is described.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 26, 2026, 01:22 AM
Security Audit — agent-trust-hub — prove-a-pr