github-workflow-automation

Pass

Audited by Gen Agent Trust Hub on Mar 26, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches and executes specialized packages such as ruv-swarm and claude-flow@alpha from the npm registry using npx. These are vendor-provided tools essential for the skill's intended functionality.
  • [REMOTE_CODE_EXECUTION]: Employs npx to execute remote scripts and tools for repository orchestration, CI/CD optimization, and self-healing pipeline management.
  • [COMMAND_EXECUTION]: Operates the GitHub CLI (gh) and git to perform repository actions, including creating issues, managing pull requests, and viewing workflow logs.
  • [DATA_EXFILTRATION]: No unauthorized data exfiltration patterns were identified. Network activity is confined to standard GitHub API interactions via official tooling and package retrieval from well-known registries.
  • [INDIRECT_PROMPT_INJECTION]: The skill features a surface for indirect prompt injection as it processes external data from pull request metadata and CI/CD run logs (e.g., via gh run view and gh pr view).
  • Ingestion points: Pull request files/labels and workflow run conclusions in SKILL.md.
  • Boundary markers: Absent from the shell piping examples.
  • Capability inventory: npx ruv-swarm provides capabilities for file modification, PR creation, and issue management.
  • Sanitization: Not explicitly detailed in the command examples, but the risk is assessed as low given the specialized nature of the processing tools.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 26, 2026, 08:45 PM