github-project-management
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill uses
npxto download and executeruv-swarmandclaude-flowpackages from the NPM registry. These are external dependencies required for the swarm coordination features. It also references theruvnet/swarm-action@v1GitHub Action in the workflow examples. - [COMMAND_EXECUTION]: The skill makes extensive use of shell commands through the GitHub CLI (
gh) andnpxto automate project management tasks, such as creating issues, managing project boards, and tracking milestones. These commands often involve subshells and piped data (e.g.,gh issue view ... | jq). - [INDIRECT_PROMPT_INJECTION]:
- Ingestion points: The skill retrieves potentially untrusted content from external GitHub issues using commands like
gh issue viewandgh issue list(e.g., in the task decomposition and triage workflows inreferences/command-reference.md). - Boundary markers: The instructions specifically highlight the risk of command injection in GitHub Actions and recommend using environment variables instead of inlining context, but there are no explicit delimiters defined for the prompt interpolation within the swarm coordination tools themselves.
- Capability inventory: The skill possesses significant capabilities including creating/editing issues, posting comments, and modifying GitHub Project boards across repositories.
- Sanitization: There is no evidence of explicit sanitization or filtering of issue body content or comments before they are processed by the AI swarm tools.
Audit Metadata