review-and-commit
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data such as Git diffs and repository file content during its review workflow, creating a potential surface for indirect prompt injection.
- Ingestion points: Git diff outputs and repository source files as described in both
SKILL.mdandreferences/current-guidance.md. - Boundary markers: The skill requires the use of structured evidence blocks (e.g.,
owner:,allowed imports:) which help categorize review findings and separate them from the code being analyzed. - Capability inventory: The skill is authorized to perform Git operations (
git merge,git write) and execute internal validation scripts (moru-hook,agent-structure-check.py). - Sanitization: No explicit sanitization or escaping of the diff content is mentioned; the skill relies on the structured "Review Hook" for validation.
- [COMMAND_EXECUTION]: The skill instructs the agent to use standard Git commands and project-specific internal utilities to perform code reviews and manage commits. These operations are essential to the skill's primary function and are performed using local paths and placeholders.
Audit Metadata