mo-review
Pass
Audited by Gen Agent Trust Hub on Aug 12, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input data from the local environment, including Git diffs and project documentation, which could contain instructions intended to influence the agent's review logic. 1. Ingestion points: Reads from docs/business.md, docs/business/index.md, and Git diffs. 2. Boundary markers: The skill does not define specific delimiters to isolate untrusted content from its instructions. 3. Capability inventory: The agent can launch subagents and modify files to apply findings. 4. Sanitization: No specific sanitization or filtering logic is described for the processed documentation.
- [COMMAND_EXECUTION]: The skill references the execution of local scripts and tools such as mo-models.mjs to determine model routing and availability. These tools are part of the broader development environment provided by the author.
Audit Metadata