github-release-management

Warn

Audited by Gen Agent Trust Hub on Sep 18, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill frequently invokes npx claude-flow and npx claude-flow@alpha in SKILL.md to execute release management and swarm coordination tasks. These commands fetch and run code from the npm registry at runtime, introducing a dependency on an external package whose security profile should be reviewed.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data retrieved from GitHub, creating a potential surface for indirect prompt injection.
  • Ingestion points: The skill fetches commit messages and Pull Request metadata (titles, labels, authors) from GitHub via gh api and gh pr list commands in SKILL.md, storing them in variables for processing.
  • Boundary markers: The instructions in SKILL.md do not include explicit delimiters or warnings to the agent regarding the untrusted nature of these external inputs.
  • Capability inventory: As documented in SKILL.md, the skill utilizes powerful tools including shell command execution (Bash), file system modifications (Write, Edit), and network-based deployment operations (npm publish, gh release upload, docker push).
  • Sanitization: There are no documented steps in SKILL.md for validating or sanitizing the GitHub metadata before it is interpolated into commands or processed by AI agents.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Sep 18, 2026, 06:15 PM
Security Audit — agent-trust-hub — github-release-management