MCP endpoint for AI agents
Connect an AI agent to the socket.help MCP endpoint at /mcp to run utilities and manage webhook inboxes, mock APIs, files and mock MCP servers over HTTP.
Updated 27 October 2026What the platform MCP endpoint does
https://socket.help/mcp is a Model Context Protocol server that gives an agent the same capabilities as the HTTP API: run any executable utility, and create, inspect, renew, share and delete temporary resources. It shares permissions and quotas with HTTP, so an agent using your API key works within your workspace plan.
This endpoint is different from a mock MCP server. The platform endpoint exposes socket.help itself; a mock MCP server at /mcp/mock/{id} returns tools and responses you configure.
Connect a client
Configure your client for Streamable HTTP with the URL https://socket.help/mcp. Add Authorization: Bearer with an API key to use your workspace; without it the agent works anonymously. Initialize before calling tools.
Requests are POST with Accept: application/json, text/event-stream. Responses are plain JSON. GET returns 405 because SSE streams and transport sessions are not provided. Supported protocol versions are 2025-11-25, 2025-06-18 and 2025-03-26. Platform request bodies can be up to 16 MiB, including Base64 file input, with four concurrent requests admitted before body parsing. Individual file limits still apply.
curl https://socket.help/mcp \
-H "Authorization: Bearer $SOCKET_HELP_API_KEY" \
-H 'Accept: application/json, text/event-stream' \
-H 'Content-Type: application/json' \
-d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{
"name":"run_tool",
"arguments":{"toolId":"json","input":"{\"hello\":true}"}
}}'Available tools
tools/list returns these tools with input schemas and annotations. Tool results are text content containing JSON.
| Tool | Purpose |
|---|---|
| list_tools | List executable utility IDs |
| describe_tool | Options, defaults, limits and examples for one tool ID |
| run_tool | Run a utility with toolId, input, options and file or files (1 to 20) |
| create_webhook_inbox | Create an inbox; optional visibility |
| inspect_webhook_inbox | Read inbox details and captured requests |
| delete_webhook_inbox | Delete an inbox and its history |
| create_mock_api | Create a mock API from name and routes |
| inspect_mock_api | Read a mock configuration and usage |
| delete_mock_api | Delete a mock API |
| upload_temporary_file | Upload a file up to 64 KiB as Base64 |
| inspect_temporary_file | Read file metadata and remaining downloads |
| download_temporary_file | Download a file up to 64 KiB as Base64; uses one download |
| delete_temporary_file | Delete a file |
| resource_request | Inspect, update, delete, renew, change sharing or list resources of any kind; create inboxes, mocks and mock MCP servers |
Credentials inside tool calls
The inspect_* and delete_* tools require id and token: the inspection or management token returned at creation. For workspace resources created with the connection’s API key, resource_request works without a token. download_temporary_file takes the private access token as accessToken, separate from the management token.
resource_request takes kind (inboxes, mocks, files or mcp-servers), action (create, inspect, update, delete, renew, sharing or list), and optionally id, token, config, ttlSeconds, persistent and visibility. It calls the same HTTP endpoints, so its results match the HTTP responses.
Errors
Invalid arguments return a JSON-RPC error with code -32602, and unknown methods return -32601. API errors such as quota or permission failures come back as a tool result with isError: true and a JSON body containing error and status.
Frequently asked questions
What is the socket.help MCP server URL?
https://socket.help/mcp, using Streamable HTTP. Send Authorization: Bearer with an API key to use your workspace allowances.
Can an agent use socket.help MCP without an account?
Yes. Without Authorization it works with anonymous allowances, and resources it creates are managed with the tokens returned in the tool results.
Why does my MCP client get 406 or 405?
406 means the Accept header does not include both application/json and text/event-stream. 405 means the client sent GET to open an SSE stream, which this endpoint does not provide.
How does an agent discover which utilities it can run?
Call list_tools for the executable IDs, then describe_tool for a tool’s options, defaults and limits before calling run_tool.