MCP bridge

Consume or publish Model Context Protocol tools. Interop with Claude Desktop, Cursor, Continue, etc.

#Consume an MCP server

import { createMcpClient, toolsFromMcpClient } from '@agentskit/tools'

const client = await createMcpClient({
  transport: 'stdio',
  command: 'my-mcp-server',
})

const mcpTools = await toolsFromMcpClient(client)

const runtime = createRuntime({ adapter, tools: [...mcpTools, ...myTools] })

#Publish AgentsKit tools as MCP

import { createMcpServer } from '@agentskit/tools'

const server = createMcpServer({
  name: 'agentskit-devtools',
  version: '0.1.0',
  tools: [github(...), slack(...)],
})

await server.listen({ transport: 'stdio' })

#Transports

stdio — sub-process pipe
http — HTTP/SSE
ws — WebSocket

#Devtools over MCP

@agentskit/tools/mcp-devtools exposes a running runtime as MCP tools — any MCP-aware client (Claude Code, Cursor, Codex) can list sessions, inspect messages, pause / step, replay, and run evals.

import { createMcpServer } from '@agentskit/tools'
import { devtoolsTools } from '@agentskit/tools/mcp-devtools'

const server = createMcpServer({
  name: 'agentskit-devtools',
  version: '0.1.0',
  tools: devtoolsTools({ inspector: myRuntimeInspector }),
})

await server.listen({ transport: 'stdio' })

The inspector is a capability bag — only methods you implement get exposed. Read-only consumers pass listSessions + inspectSession and never expose pause / step / replay. Auth lives at the transport (bearer header on HTTP/WS, file perms on stdio).

Explore nearby

✎ Edit this page on GitHub·Found a problem? Open an issue →·How to contribute →