triaging-issues
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the GitHub CLI (
gh) to perform repository analysis. It includes a dedicated constraints section that explicitly prohibits modification actions such as editing, closing, or labeling issues, and merging or reviewing pull requests, ensuring the agent operates in a read-only capacity. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data by reading GitHub issue and pull request content, which presents a surface for indirect prompt injection.
- Ingestion points: Issue and pull request titles, bodies, and diffs accessed via
gh issue view,gh pr view, andgh pr diffcommands. - Boundary markers: The instructions do not define specific delimiters or instructions for the agent to ignore potentially malicious instructions embedded within the issue content.
- Capability inventory: The skill's capabilities are limited to read-only GitHub CLI commands; it does not have the ability to write to the filesystem or perform arbitrary network requests.
- Sanitization: No evidence of sanitization or filtering of the ingested issue or PR data is provided in the instructions.
Audit Metadata