Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.corsair.dev/llms.txt

Use this file to discover all available pages before exploring further.

Agent SDKs use the same hosted MCP server as coding agents. If you already run Vercel AI SDK, OpenAI Agents, or Claude Agent SDK, point them at Corsair’s MCP endpoint. In the dashboard: create an instance, add a plugin, add a tenant, connect credentials, then open the tenant side panel for the MCP URL and API keys. Quick Start is an optional guided demo.

​
Quick example (Vercel AI SDK)

When your backend already has the developer API key, use tenant.mcp.createVercelClient() — no separate MCP secret required: 
import { openai } from "@ai-sdk/openai";
import { createClient } from "@corsair-dev/app";
import { stepCountIs, streamText } from "ai";

const corsair = createClient({ apiKey: process.env.CORSAIR_DEV_KEY! });
const inst = corsair.instance(process.env.CORSAIR_INSTANCE_ID!);
const mcpClient = await inst.tenant("alice").mcp.createVercelClient();

const result = streamText({
  model: openai("gpt-4.1"),
  tools: await mcpClient.tools(),
  stopWhen: stepCountIs(15),
  prompt: message,
});
Install peers: 
npm install @corsair-dev/app ai @ai-sdk/mcp @ai-sdk/openai

​
Alternative: tenant MCP key

If you prefer a dedicated bearer token per agent (e.g. for a user-facing connector), create a key and pass url + apiKey to the adapter helpers: 
import { createVercelAiMcpClient } from "@corsair-dev/app";

const key = await inst.tenant("alice").mcpKeys.create("vercel-agent");

const mcpClient = await createVercelAiMcpClient({
  url: key.mcpHttpUrl,
  apiKey: key.secret,
});

​
Framework guides

Vercel AI SDK

createVercelClient() or createVercelAiMcpClient()

OpenAI

Responses API and OpenAI Agents SDK

Claude

Claude Agent SDK mcpServers

Custom connector

Any HTTP MCP client