DEVELOPERS

Open to AI and automation

Two ways in: the REST API plugs contacts, deals, tasks and your knowledge base into your own workflows; MCP lets Claude, ChatGPT and Cursor read — and safely write to — your CRM. All endpoints are live in production.

GitHub repo Start Personal Trial

MCP

Connect your AI assistant to your CRM

SiliMind speaks MCP (Model Context Protocol) — the standard way for AI assistants to use external tools. Your assistant can directly read (and, with your approval, write to) your SiliMind CRM.

What it connects to

🤖

Claude Desktop

Ask Claude about your deals, customers, tasks

💬

ChatGPT

Query your pipeline through custom GPTs

⚙️

Cursor / any MCP client

Build your own AI workflows on SiliMind data

What your AI can do

🔍 Search customers	“Find all contacts at Acme Corp”
📊 Check deals	“What’s in my follow-up pipeline?”
📅 View calendar	“What’s on my schedule tomorrow?”
📝 Create tasks	“Add a follow-up task for John” (requires your approval in SiliMind)
🧠 Query anything	Natural language questions about your allowed data

Connect in three steps

STEP 1 · Get your token

SiliMind → Settings → MCP Access → Generate.

STEP 2 · Add to Claude Desktop

Add this to your Claude Desktop config:

{
  "mcpServers": {
    "silimind": {
      "command": "python",
      "args": ["-m", "silimind_mcp.stdio_bridge"],
      "env": {
        "SILIMIND_URL": "https://agent.silimind.net",
        "SILIMIND_TOKEN": "smcp_your_token"
      }
    }
  }
}

STEP 3 · Ask away

“Hey Claude, what are my top deals this week?”

Security first

🔒 Per-token permissionsreadonly / readwrite granularity

🔒 Tenant isolationyour AI sees only your data

🔒 Rate-limited, fully audit-loggedevery call is traceable

🔒 Writes need your approvalAI proposes, you approve inside SiliMind

API

REST API

All endpoints use Bearer token auth and cover core-resource CRUD, AI suggestions and knowledge-base semantic search.

Base URL: https://agent.silimind.net/api

Authentication

POST /api/auth/login
Content-Type: application/json

{
  "email": "you@company.com",
  "password": "your_password"
}

Response:

{
  "accessToken": "eyJhbGciOiJIUzI1NiIs...",
  "tokenType": "bearer",
  "role": "sales"
}

Use the token in all subsequent requests:

Authorization: Bearer eyJhbG…s...

Core resources

Contacts

Method	Endpoint	Description
GET	/contacts	List contacts (paginated, searchable)
POST	/contacts	Create a contact
GET	/contacts/{id}	Get contact detail
PUT	/contacts/{id}	Update a contact
DELETE	/contacts/{id}	Delete a contact

Companies

Method	Endpoint	Description
GET	/companies	List companies
POST	/companies	Create a company
GET	/companies/{id}	Get company detail
PUT	/companies/{id}	Update
DELETE	/companies/{id}	Delete

Deals

Method	Endpoint	Description
GET	/deals	List deals
POST	/deals	Create a deal
GET	/deals/{id}	Get deal detail
PUT	/deals/{id}	Update deal (stage, amount, etc.)
DELETE	/deals/{id}	Delete

Tasks

Method	Endpoint	Description
GET	/tasks	List tasks
POST	/tasks	Create a task
PUT	/tasks/{id}	Update (complete, reschedule)
DELETE	/tasks/{id}	Delete

Interactions

Method	Endpoint	Description
GET	/interactions	List interaction records
POST	/interactions	Log an interaction (call, email, meeting)

Visits

Method	Endpoint	Description
GET	/visits	List visits
POST	/visits	Schedule a visit

Example — search contacts

GET /api/contacts?q=acme&page=1&page_size=20
Authorization: Bearer <your-token>

Example — create a contact

POST /api/contacts
Content-Type: application/json
Authorization: Bearer <your-token>

{
  "name": "John Smith",
  "email": "john@acme.com",
  "company": "Acme Corp",
  "phone": "+86 138 0000 0000"
}

Example — list your pipeline

GET /api/deals?stage=follow_up
Authorization: Bearer <your-token>

AI features

AI suggestions

GET /api/ai/suggestions
Authorization: Bearer <your-token>

AI chat (agent)

POST /api/agent-chat/{agent_id}/message
Content-Type: application/json
Authorization: Bearer <your-token>

{"content": "帮我总结这个客户最近的沟通"}

Streaming SSE response.

Knowledge base

Method	Endpoint	Description
GET	/knowledge/docs	List documents
POST	/knowledge/docs/upload	Upload document (PDF/DOCX/MD)
POST	/knowledge/search	Semantic search

Example — semantic search

POST /api/knowledge/search
Content-Type: application/json
Authorization: Bearer <your-token>

{"query": "产品规格参数", "top_k": 5}

Response conventions

Success — 200 / 201 (created)
Auth failure — 401 — token expired or invalid
Permission denied — 403
Not found — 404
Validation error — 422 with field details
Rate limit — 429

Pagination

List endpoints use page and page_size query parameters (default: 20, max: 100).

Rate limits

Per-token rate limiting applies
Exceeding limits returns 429 with a Retry-After header

Support

Docs: https://www.silimind.net · MCP integration: github.com/Leonchen3352/silimind-mcp · Contact: micheal@silimind.net

Browse the MCP integration on GitHub

Client source, setup examples and issue tracker.

Open the silimind-mcp repo →