github-actions-author
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes git and the GitHub CLI (gh) to list workflows, check repository status, and fetch release information from well-known repositories such as actions/checkout and oven-sh/setup-bun.
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data including workflow files and project manifests. * Ingestion points: Existing workflow YAML files, package.json, and lockfiles (SKILL.md, steps 1 & 2). * Boundary markers: Instructions lack explicit text delimiters for user content but emphasize safety warnings and secure handling practices. * Capability inventory: The skill uses git, gh, and bun to read and write repository content (SKILL.md, steps 1, 4, 6). * Sanitization: Instructions mandate environment variable usage for untrusted event data and prohibit direct interpolation of PR/issue text (SKILL.md, step 5 & Security Review Checklist).
Audit Metadata