to-issues-clearly
Pass
Audited by Gen Agent Trust Hub on Nov 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the GitHub CLI (
gh) and the GitHub API to automate the creation, modification, and organization of issues and milestones. It also calls local Python utility scripts (audit_issues.pyandblockers.py) to verify issue integrity and manage dependencies. - [INDIRECT_PROMPT_INJECTION]: The skill possesses a data ingestion surface as it processes external, potentially untrusted source materials (PRDs, plans, bug reports). However, the risk is addressed through a structured translation process.
- Ingestion points: Processes external source materials, including plans and existing repository issues.
- Boundary markers: The skill employs explicit markdown templates (e.g.,
## Outcome,## Done when) to structure outputs and separate processed content. - Capability inventory: Includes shell command execution via
ghCLI and local Python scripts, as well as GitHub API interaction. - Sanitization: The instructions mandate a 'verifier-language rewrite pass', requiring the agent to translate source concepts into domain-specific or code-internal language, which functions as a form of semantic sanitization.
- [EXTERNAL_DOWNLOADS]: The skill interacts with GitHub's official API to fetch metadata such as milestones and existing issues. These operations target a well-known service and are consistent with the skill's stated administrative purpose.
Audit Metadata