MCP Tool Permission Planner. Allow, ask or block, tool by tool.
Turn a server's tool list into a clear permission policy. Paste tool names or a tools/list response, check each tool's type and copy the table.
Plan your tool permissions
Paste tool names or JSON. It is read in your browser and sent nowhere.
Annotations such as readOnlyHint and destructiveHint are defined in the MCP specification, which says clients must treat them as untrusted unless the server is trusted. See MCP tool annotations.
How to use it
How to use the MCP Tool Permission Planner
Paste one tool name per line, or the JSON from a tools/list response, for example from the MCP Inspector. The planner pre-fills each tool's type from its annotations or, failing that, its name.
Check every row and change the type where needed. Reads are auto-allowed, writes ask every time and destructive tools are blocked. Tick the sensitive data box to ask before reads too, then copy the policy as a table or a list.
Limits
What it can't tell you
It doesn't connect to the server or enforce anything. You still set the permissions in your AI app or gateway. The pre-filled types are guesses from names and hints.
Some risks come from combinations, not single tools. A read tool and a send tool connected at the same time can leak data together. See MCP data exfiltration.
FAQ
Common questions
What counts as a destructive MCP tool?
Should read-only tools ask for approval?
Where do I set these permissions?
How do I get a server's tool list?
Project Managers who will work with you on your project!
Irina S.
Benjamin G.
David G.
Jeanette M.
Mark S.
Maryna P.
Wondering what this costs? Every service has published pricing, and you don't need a discovery call to see it.
View pricingWant a second opinion on your MCP setup?
Request a free MCP readiness review. We'll look at the tools and AI apps you use and tell you where MCP could help, which servers exist and what to lock down first.
Request a Free Review →
