AI agents use refresh_site to create or update resources in Docs — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Docs environment.
This tool fetches external content and updates the local SQLite index with changed pages. It modifies stored data (writing updated page content to the database) but does so reversibly — the index can be rebuilt or updated again. It is not purely a Read operation because it causes side effects (database writes), and it does not irreversibly destroy data, execute arbitrary code, or involve financial transactions.
From the tool's definition Re-crawls a site, updating only changed pages by default
Attacks that exploit this kind of access
Re-crawls a site, updating only changed pages by default. It is categorised as a Write tool in the Docs MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Docs MCP server in PolicyLayer and add a rule for refresh_site: 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.
refresh_site is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the refresh_site 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 refresh_site. 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.
refresh_site 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.
refresh_site is one line of Docs's registry record.
The record carries the whole server: verified identity, auth posture, risk grade, every tool classified, recommended policy — re-checked continuously.
Teams ship this data inside their own products. See what a licence covers →