contributors-update

Pass

Audited by Gen Agent Trust Hub on Aug 23, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill identifies a potential surface for indirect prompt injection by processing external metadata from pull requests.\n
  • Ingestion points: Pull request titles and descriptions retrieved via the GitHub CLI (SKILL.md).\n
  • Boundary markers: Step 4 explicitly instructs the agent to treat contributor-authored metadata as untrusted and to ignore any embedded instructions.\n
  • Capability inventory: The skill possesses the capability to execute shell commands and perform remote repository updates (git push).\n
  • Sanitization: The skill relies on narrative warnings to the agent and a mandatory human review step rather than automated data sanitization.\n- [COMMAND_EXECUTION]: The skill utilizes shell commands to manage the local environment and interact with the remote repository.\n
  • Evidence: Execution of git commands for branch management, synchronization, and commits in Step 1 and Step 6.\n
  • Evidence: Use of the GitHub CLI (gh) to fetch repository metadata in Step 3.\n
  • Risk reduction: The skill implements a safety check to verify that only the intended file (README.md) is modified before finalizing the commit.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 23, 2026, 03:51 PM
Security Audit — agent-trust-hub — contributors-update