Search the text contents of a previously cloned repository (by
AI agents call repo_grep to retrieve information from Kubesearch without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool performs text searching on already-cloned repository contents. It retrieves and queries data without modifying, deleting, or executing anything. The 'previously cloned' qualifier confirms the repository is already present, and grep only searches—it does not create, modify, or delete files. This is a classic Read operation with minimal blast radius even if misused.
From the tool's definition Tool name 'repo_grep' and description 'Search the text contents of a previously cloned repository' indicate a read-only search operation. Grep is a standard text search utility with no side effects.
Attacks that exploit this kind of access
Search the text contents of a previously cloned repository (by. It is categorised as a Read tool in the Kubesearch MCP Server, which means it retrieves data without modifying state.
Register the Kubesearch MCP server in PolicyLayer and add a rule for repo_grep: 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 Kubesearch. Nothing to install.
repo_grep 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 repo_grep 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 repo_grep. 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.
repo_grep is provided by the Kubesearch MCP server (perfectra1n/kubesearch-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 →