coding-review-simplify
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of code diffs and file content during its review workflow. This data is used to drive the behavior of three sub-agents and can influence file modifications or shell command execution.
- Ingestion points: The skill reads the user request, changed files, and code diffs as primary inputs for analysis (
SKILL.md). - Boundary markers: The instructions lack explicit boundary markers or instructions to isolate untrusted code content from the agent's logic, increasing the risk of the agent following instructions embedded within the code being reviewed.
- Capability inventory: The skill uses high-privilege capabilities including shell execution (
Bash), file modification (Edit), and sub-task delegation (Agent) across its scripts. - Sanitization: There is no explicit sanitization, validation, or filtering of the input code diffs before they are processed by the sub-agents or used to suggest and apply code changes.
Audit Metadata