repo-admin

Pass

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill performs sensitive administrative operations on GitHub repositories, such as managing collaborators, updating visibility, and configuring branch protection. These high-privilege actions are mitigated by a mandatory workflow involving user previews and explicit confirmation.
  • [PROMPT_INJECTION]: The skill processes untrusted data from GitHub API responses (such as label names and milestone titles) which could serve as a vector for indirect prompt injection. 1. Ingestion points: GitHub API responses fetched via MCP tools (SKILL.md). 2. Boundary markers: Not explicitly defined for processed strings. 3. Capability inventory: High-impact administrative tools like collaborator removal (SKILL.md). 4. Sanitization: Safety relies on mandatory user confirmation before performing actions.
  • [SAFE]: All operations rely on official GitHub API tools and documentation. There are no external dependencies, remote script executions, or obfuscated contents.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 10, 2026, 05:34 AM
Security Audit — agent-trust-hub — repo-admin