qe-github-release-management

Pass

Audited by Gen Agent Trust Hub on Feb 27, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill relies on the external package 'claude-flow', which is downloaded and executed at runtime via 'npx'.
  • Evidence: Multiple occurrences of 'npx claude-flow' and 'npx claude-flow@alpha' throughout SKILL.md for release management, changelog generation, and swarm initialization.
  • [COMMAND_EXECUTION]: The skill frequently executes shell commands to interact with the GitHub CLI (gh), git, and npm.
  • Evidence: Use of the 'Bash' tool to run 'gh api', 'gh release', 'npm version', and 'git checkout'.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it ingests untrusted external data that is then processed by AI agents.
  • Ingestion points: Git commit messages and Pull Request titles/labels/bodies fetched via 'gh api' and 'gh pr list' in SKILL.md.
  • Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present when passing this data to the 'claude-flow' tool.
  • Capability inventory: The skill possesses significant capabilities including file writing ('Write'), arbitrary shell command execution ('Bash'), and the ability to spawn further agents ('mcp__claude-flow__agent_spawn').
  • Sanitization: No visible evidence of escaping or validating the content of commits or PR descriptions before they are used to generate changelogs or influence release logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 27, 2026, 06:09 PM