Search the local knowledge base for authoritative documentation.
AI agents call grounded_search to retrieve information from Grounded Code MCP without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool searches a local knowledge base, which is a read operation with no side effects. It retrieves data from a pre-existing collection of books, standards, and documentation. There is no indication it can modify, delete, execute code, or perform financial operations. The blast radius of misuse is minimal—worst case, an AI agent returns irrelevant or incorrect search results from available documentation.
From the tool's definition Tool description states 'Search the local knowledge base for authoritative documentation.' The name 'grounded_search' and verb 'search' indicate a retrieval operation.
Attacks that exploit this kind of access
Search the local knowledge base for authoritative documentation. It is categorised as a Read tool in the Grounded Code MCP MCP Server, which means it retrieves data without modifying state.
Register the Grounded Code MCP server in PolicyLayer and add a rule for grounded_search: 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 Grounded Code MCP. Nothing to install.
grounded_search 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 grounded_search 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 grounded_search. 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.
grounded_search is provided by the Grounded Code MCP server (michaelalber/grounded-code-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 →