About Typeship
Typeship’s MCP server lets AI agents inspect and operate the Typeship API for resolving and diagnosing OpenAPI or GraphQL specs and maintaining selected SDK, CLI, and MCP targets.
Key features
- 01Search and read documentation
- 02Execute API operations
- 03Validate typed operation arguments
- 04Restrict returned fields
- 05Read-only mode and tool selection
Use cases
- 01Find API guidance before acting
- 02Resolve or diagnose an API specification
- 03Maintain selected generated targets
MCP support
Exposes Typeship API documentation and operations to MCP clients through discovery tools and execute. Supports local stdio and hosted connections, with optional read-only mode.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://typeship.dev/mcpclaude mcp add --transport http dev-typeship-typeship https://typeship.dev/mcp{"mcpServers":{"dev-typeship-typeship":{"type":"http","url":"https://typeship.dev/mcp"}}}Open in Cursor{"servers":{"dev-typeship-typeship":{"type":"http","url":"https://typeship.dev/mcp"}}}Open in VS Code{"mcpServers":{"dev-typeship-typeship":{"type":"http","url":"https://typeship.dev/mcp"}}}{"mcpServers":{"dev-typeship-typeship":{"serverUrl":"https://typeship.dev/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
https://typeship.dev/mcphttps://typeship.dev/mcp/publichttps://typeship.dev/mcp/readonlyTools
Details
From the project README@typeship-ax/mcp
MCP server for the Typeship API. API reference
Resolve an OpenAPI or GraphQL Spec, diagnose it, and keep every selected CLI, MCP, and SDK Target current.
Installation
MCP clients start the server with npx, so it needs no separate installation (see Connect an MCP client). To install the typeship-mcp command globally instead:
npm install --global @typeship-ax/[email protected]
Requires Node.js 20+.
Connect an MCP client
Provide TYPESHIP_TOKEN through the MCP client's environment or secret settings. Keep credential values out of URLs and command arguments.
Local
- Claude Code:
claude mcp add typeship -- npx -y --package @typeship-ax/mcp typeship-mcp - Codex:
codex mcp add typeship -- npx -y --package @typeship-ax/mcp typeship-mcp - Install in VS Code
Local · read-only
- Claude Code:
claude mcp add typeship-readonly -- npx -y --package @typeship-ax/mcp typeship-mcp --read-only - Codex:
codex mcp add typeship-readonly -- npx -y --package @typeship-ax/mcp typeship-mcp --read-only - Install in VS Code
Hosted
- Claude Code:
claude mcp add --transport http typeship https://typeship.dev/mcp - Codex:
codex mcp add typeship --url https://typeship.dev/mcp - Install in VS Code
Hosted · read-only
- Claude Code:
claude mcp add --transport http typeship-readonly https://typeship.dev/mcp/readonly - Codex:
codex mcp add typeship-readonly --url https://typeship.dev/mcp/readonly - Install in VS Code
Other clients
Add a server entry to your client's MCP configuration. For Cursor, merge it into mcpServers in .cursor/mcp.json, then enable the server in Cursor's MCP settings:
{
"mcpServers": {
"typeship": {
"command": "npx",
"args": [
"-y",
"--package",
"@typeship-ax/mcp",
"typeship-mcp"
],
"env": {
"TYPESHIP_TOKEN": "replace-with-your-credential"
}
}
}
}
Replace the credential placeholder using the MCP client's secret storage when it has one. The server reads TYPESHIP_TOKEN from its environment; credentials never belong in command arguments.
The server supports MCP 2025-11-25 and 2026-07-28 and picks the version automatically, so clients need no protocol settings. After registering it with Claude Code, claude mcp list shows the connection.
Tools
The server runs over stdio with no runtime dependencies and exposes a compact discovery surface: search_docs, read_docs, and execute. Read an operation before executing it to get its arguments, an example, and its safety classification.
Tool input schemas come from the OpenAPI spec, so agents see real parameter types and required fields. Arguments are checked before anything reaches the API (unknown or mistyped ones come back as one isError result, nothing is dropped), every tool takes fields to keep only the result keys it needs, and errors carry a stable code and next_steps.
Add --read-only to args (or set TYPESHIP_MCP_READ_ONLY=1) for a server that cannot write, --tools projects,specs (or TYPESHIP_MCP_TOOLS) to expose a subset, and TYPESHIP_MCP_MAX_RESULT_CHARS to change the result size cap (64,000).
Generated from the OpenAPI spec by Typeship.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.