github-actions

Pass

Audited by Gen Agent Trust Hub on Sep 25, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions to download the official GitHub Actions runner package from the GitHub releases page. This targets a well-known service and official repository, which is a standard procedure for infrastructure setup.
  • [INDIRECT_PROMPT_INJECTION]: The skill manages GitHub Actions workflow files (YAML), which contain instructions for automated code execution. This creates a surface where an agent could be manipulated through external repository content to modify workflows maliciously.
  • Ingestion points: Workflow files in the .github/workflows/ directory.
  • Boundary markers: None present; the skill treats the workflow files as standard configuration data.
  • Capability inventory: The skill facilitates shell command execution (via run steps), network operations (via curl), and access to sensitive repository secrets (secrets.*).
  • Sanitization: None; instructions assume the integrity of the target repository and the user's workflow definitions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 25, 2026, 12:20 AM
Security Audit — agent-trust-hub — github-actions