upgrade-github-actions

Pass

Audited by Gen Agent Trust Hub on Jul 14, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches release metadata from the official GitHub API (api.github.com). This is a well-known service, and the activity is limited to retrieving public version tags to identify available updates.
  • [COMMAND_EXECUTION]: The skill executes a bundled script for repository discovery and interacts with the git CLI. Safety is maintained through human-in-the-loop requirements, ensuring no files are changed or pushed without user approval.
  • [PROMPT_INJECTION]: The skill processes workflow files and external API responses, which constitutes an ingestion surface for indirect prompt injection. Ingestion points: Workflow files in the .github/workflows/ directory and JSON data from api.github.com. Boundary markers: The skill relies on agent instructions for human verification rather than explicit delimiters in script output. Capability inventory: The skill can modify repository files and execute git commands. Sanitization: Discovery is performed using specific regex patterns, and the agent is instructed to manually audit release notes for potential breaking changes.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 14, 2026, 12:00 PM
Security Audit — agent-trust-hub — upgrade-github-actions