See every tool
an MCP server exposes.
Enter a domain or MCP endpoint to list its tools, the parameters each one takes, which arguments are required, and whether a tool reads, changes or deletes data.
Every result is a full report with the server's tools, parameters and a score out of 100.
What you see
Know what a server can do before you give it to an agent
Every tool and function
The full tools/list response, paged through to the end, with each tool's name, title and description.
Parameters and types
Each argument with its type, allowed values, default and whether it is required, read straight from the input schema.
Behavior hints
Read only, destructive, idempotent and open world hints, so you can spot tools that change or delete data.
Try it in chat
Open a live chat against the server and ask it to run a tool with a realistic example before you install it.
How it works
Three steps, about 30 seconds
- 01
Enter a server
A domain or a full endpoint. We try common paths such as /mcp when you give a domain.
- 02
We list its tools
A live MCP session pages through tools/list and reads every input and output schema.
- 03
Review before you connect
Check parameters and behavior hints, then try the tools in chat or add the server to your client.
Example reports
Servers we have already checked
Every tool you connect is a tool an agent can call.
FAQ
Questions, answered
How do I see what tools an MCP server has?
Enter the server's endpoint or domain above. We connect with the Model Context Protocol, call tools/list and show every tool with its description, parameters and behavior hints on the report page.
Are MCP tools the same as functions?
Yes. An MCP tool is a function the server lets an AI model call. Each one has a name, a description and a JSON Schema for its arguments, which the model uses to decide when and how to call it.
What do the behavior hints mean?
Read only means the tool does not change anything. Destructive means it may delete or overwrite data. Idempotent means repeating a call has no extra effect. Open world means it reaches outside systems such as the web.
Can I see resources and prompts too?
The report shows how many resources and prompts the server offers alongside its tools. Tools get the full breakdown because they are what agents call.
Can I list tools on a server that needs auth?
Yes. Add an Authorization header under the form. It is used for that one session and never stored.
Does listing tools run any of them?
No. Listing only reads tool definitions. Nothing is called unless you open the chat and ask the server to run a tool.
See every tool
an MCP server exposes.
Free, no signup, results in under 30 seconds.
List tools