github-actions-hardening
Pass
Audited by Gen Agent Trust Hub on Aug 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides comprehensive security auditing instructions focused on hardening GitHub Actions workflows. It correctly identifies and provides remediation for high-risk configurations, such as shell injection via interpolation, over-privileged GITHUB_TOKEN scopes, and the dangers of the pull_request_target trigger when checking out fork code. The guidance follows industry best practices, such as SHA-pinning actions and using OpenID Connect (OIDC) for cloud authentication.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze untrusted GitHub Actions workflow files (.github/workflows/*.yml), creating an inherent attack surface for indirect prompt injection. 1. Ingestion points: User-provided workflow YAML files. 2. Boundary markers: Absent. 3. Capability inventory: Instructional only; no active tools, file-writing, or network capabilities are requested or used. 4. Sanitization: The instructions do not specify input validation or sanitization, but they mandate a rigid, structured reporting workflow (mapping triggers, hunting sinks, auditing permissions) which acts as a framework to reduce the effectiveness of simple deceptive injections.
Audit Metadata