Let Gemini review code, architecture design, or technical solutions
AI agents call gemini_review to retrieve information from Claude Code Gemini MCP without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool sends content to Gemini for review/analysis and returns feedback. It reads/analyzes the provided content but does not modify, execute, or delete anything. The blast radius is low since it only produces a review output.
From the tool's definition 'review code, architecture design, or technical solutions' — the tool analyzes and provides feedback without modifying anything
Attacks that exploit this kind of access
Let Gemini review code, architecture design, or technical solutions. It is categorised as a Read tool in the Claude Code Gemini MCP MCP Server, which means it retrieves data without modifying state.
Register the Claude Code Gemini MCP server in PolicyLayer and add a rule for gemini_review: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Claude Code Gemini MCP. Nothing to install.
gemini_review is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the gemini_review rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for gemini_review. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
gemini_review is provided by the Claude Code Gemini MCP server (shunl12324/claude-code-gemini-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →