For developers

The same brain, wired into your own tools — over MCP or a standard API.

MCP endpoint

A Streamable HTTP MCP server at /api/mcp. Point any MCP client at it, authenticate with an agent token, and the brain tools appear in your client — no local relay to install.

Standard API

The full brain pipeline behind a chat-completions contract at /api/v1/chat/completions. Any SDK or agent framework that speaks the protocol works unchanged.

A budgeted token, not your API key

Issue an agent token with its own per-minute and per-day limits. Give the token to the agent — your account and spend stay under your control.

MCP client config

JSON

{
  "mcpServers": {
    "vesqor": {
      "type": "http",
      "url": "https://brain-shield.vercel.app/api/mcp",
      "headers": {
        "Authorization": "Bearer bs_live_..."
      }
    }
  }
}

API call

curl

curl https://brain-shield.vercel.app/api/v1/chat/completions \
  -H "Authorization: Bearer bs_live_..." \
  -H "Content-Type: application/json" \
  -d '{
    "model": "vesqor-reasoning",
    "messages": [
      { "role": "user", "content": "Our packaging line stops for 3 hours every shift..." }
    ]
  }'

Full API reference

The complete contract — endpoints, request and response shapes — is published as an OpenAPI document.

Open the OpenAPI spec

Get an agent token

Tokens are issued from your account — sign in to create one.