git-remote-release

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests and summarizes untrusted external content, which could be used to influence the agent's output.
  • Ingestion points: Pull request titles, descriptions, and commit messages from external repositories are fetched in SKILL.md (Step 3).
  • Capability inventory: The skill executes local git commands (read-only) and uses gh api or MCP tools to retrieve data.
  • Boundary markers: No specific delimiters or instructions are provided to the agent to prevent it from following instructions found within the PR bodies.
  • Sanitization: No sanitization of PR descriptions or commit messages is performed before the agent analyzes them in SKILL.md (Step 4).
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 11:32 AM
Security Audit — agent-trust-hub — git-remote-release