commit-push-pr
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted repository data which creates an inherent indirect prompt injection surface.
- Ingestion points: The skill instructs the agent to run
git diffandgit login steps 1 and 2 ofSKILL.mdto analyze code changes and commit history. - Boundary markers: The skill employs single-quoted heredocs (
cat <<'EOF') to handle text formatting safely during command assembly, but lacks specific guidelines directing the model to ignore adversarial instructions embedded within the codebase or log text. - Capability inventory: The skill has the capability to run local shell operations including
git commit,git push, andgh pr createas detailed inSKILL.md. - Sanitization: No input validation or automated filtering of the parsed repository text is performed before it is interpolated into the model's contextual understanding.
Audit Metadata