submit-pr
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes standard version control commands using
gitand the GitHub CLI (gh) to handle branch synchronization, pushing, and pull request creation. These are legitimate operations for the skill's primary purpose. - [INDIRECT_PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection by processing commit history and diffs to generate PR content.
- Ingestion points: Step 2 of
SKILL.mdingests commit messages and file diffs usinggit logandgit diff. - Boundary markers: No explicit delimiters are used when interpolating these values into the PR generation prompt.
- Capability inventory: The skill has capabilities to commit fixes, push code, and create PRs.
- Sanitization: Commit messages are not sanitized before being used in synthesis.
- Mitigation: The risk is mitigated by mandatory
CHECKPOINTprompts in Steps 3, 4, and 6, requiring explicit user approval of all generated content and automated fixes before any remote changes are made.
Audit Metadata