AI agents call get_doc to retrieve information from Docs without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool's primary function is data retrieval ('Returns the markdown body'). Although persist:true can index content, this is a reversible write operation that adds to an index rather than modifying critical state. The tool has no destructive capabilities and poses minimal risk when misused by an AI agent, as it only retrieves or optionally caches documentation.
From the tool's definition Tool returns markdown body of documentation page, retrieves data from local index or fetches on-the-fly. The optional persist parameter indexes the page but does not modify or delete existing data—it only stores additional content for future retrieval.
Attacks that exploit this kind of access
Returns the markdown body of a documentation page. Pulls from the local index when available, otherwise fetches and converts on the fly. Pass persist:true to index the fetched page into a registered site. It is categorised as a Read tool in the Docs MCP Server, which means it retrieves data without modifying state.
Register the Docs MCP server in PolicyLayer and add a rule for get_doc: 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 Docs. Nothing to install.
get_doc 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 get_doc 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 get_doc. 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.
get_doc is provided by the Docs MCP server (kage1020/docs-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 →