github
Pass
Audited by Gen Agent Trust Hub on Oct 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from GitHub notifications, issues, and pull requests. Since it includes an 'Autonomous' mode where the agent can perform actions like merging PRs, rebasing, or labeling based on the content of these items, it is vulnerable to malicious instructions embedded in PR/issue descriptions.
- Ingestion points: GitHub API data and URL-based content extraction for issues and pull requests (SKILL.md).
- Boundary markers: No specific boundary markers or 'ignore embedded instructions' warnings are present to separate untrusted data from the agent's instructions.
- Capability inventory: The agent has access to
Bash,Read, andWritetools, and is specifically authorized to perform git operations (status, rebase) and GitHub metadata changes (labels, closing issues). - Sanitization: No sanitization or validation of the ingested GitHub content is specified before the agent evaluates 'fit', 'risk', and 'next action'.
Audit Metadata