plan-to-github-issues
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided plan files to generate GitHub issues and CI configurations, creating a potential injection surface.
- Ingestion points: The agent is instructed to find and read local markdown files matching
**/plans/plan-*.md. - Boundary markers: No specific delimiters or safety instructions are provided to the agent to treat the plan content as untrusted data when interpolating it into issue bodies or workflow templates.
- Capability inventory: The skill utilizes the GitHub CLI to create labels, create issues, and write files to the repository via the GitHub Contents API.
- Sanitization: Acceptance criteria and overview content are copied verbatim from the plan files into GitHub issues, meaning any instructions embedded in the plans could be interpreted as legitimate agent tasks.
- [COMMAND_EXECUTION]: The skill relies on the GitHub CLI (
gh) to perform repository management and metadata operations. - Evidence: Uses
gh label createfor phase tracking andgh repo viewto identify the target repository. - [DYNAMIC_EXECUTION]: The skill directs the agent to dynamically generate a CI workflow configuration (
.github/workflows/ci.yml) and push it to the repository. - Evidence: Step 4 provides a template for a GitHub Actions workflow (including Postgres/Redis containers and test execution steps) which is uploaded using
gh api --method PUTwith Base64-encoded content.
Audit Metadata