AI agents call github.search-code to retrieve information from Mcp without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
Searching code on GitHub is a read-only operation that retrieves public or accessible information without modifying, executing, or deleting anything. The tool does not execute code, modify repositories, or perform financial transactions. While searching code could theoretically reveal sensitive information if someone has access to private repositories, the tool itself performs only read operations.
From the tool's definition Tool name is 'github.search-code' and description indicates it 'Search[es] code across GitHub with the code-search syntax'. This is a query/retrieval operation with no side effects.
Attacks that exploit this kind of access
Search code across GitHub with the code-search syntax (e.g. It is categorised as a Read tool in the Mcp MCP Server, which means it retrieves data without modifying state.
Register the MCP server in PolicyLayer and add a rule for github.search-code: 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 Mcp. Nothing to install.
github.search-code 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 github.search-code 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 github.search-code. 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.
github.search-code is provided by the MCP server (@2sio/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 →