MCP Server

Model Context Protocol Server

worcflow.ai exposes an MCP server so any MCP-compatible agent can call our tools directly. Query commitments, get at-risk items, and send nudges with structured inputs and outputs — no prompt engineering needed.

What is MCP?

The Model Context Protocol (MCP) is the open standard for connecting AI agents to external tools and data sources. With 1,000+ community servers and native support in Claude, OpenAI, LangChain, CrewAI, and Microsoft Copilot Studio, MCP is the standard for agent tool use in 2026.

Connecting

Add the worcflow.ai MCP server to your agent configuration.

claude_desktop_config.json

{
  "mcpServers": {
    "worcflow": {
      "url": "https://mcp.worcflow.ai/sse",
      "headers": {
        "Authorization": "Bearer agent_sk_..."
      }
    }
  }
}

Python (LangChain)

from langchain_mcp import MCPToolkit

toolkit = MCPToolkit(
    server_url="https://mcp.worcflow.ai/sse",
    headers={"Authorization": "Bearer agent_sk_..."}
)
tools = toolkit.get_tools()
# Available: query_commitments, get_at_risk, send_nudge, list_workspaces

Available Tools

query_commitments

Retrieve and filter tracked commitments by status, owner, client, date range, or confidence threshold.

Input Parameters

statusstringon_track | at_risk | overdue | completed

ownerstringFilter by commitment owner

clientstringFilter by client name

due_beforestringISO date — commitments due before this date

limitnumberMax results (default: 20)

Example Response

{
  "commitments": [
    {
      "id": "cmt_a1b2c3",
      "text": "Will ship revised samples by Thursday",
      "owner": "Rajesh",
      "status": "at_risk",
      "due": "2026-02-19T18:00:00+05:30",
      "source": "whatsapp",
      "confidence": 0.94
    }
  ],
  "total": 1
}

get_at_risk

Get commitments approaching their deadline without completion confirmation, sorted by urgency.

Input Parameters

hours_aheadnumberLook-ahead window in hours (default: 24)

Example Response

{
  "commitments": [
    {
      "id": "cmt_d4e5f6",
      "text": "Revised proposal ready by Monday EOD",
      "owner": "Priya Sharma",
      "status": "at_risk",
      "due": "2026-02-17T18:00:00+05:30",
      "source": "email"
    }
  ],
  "total": 1
}

send_nudge

Send a reminder to the commitment owner via WhatsApp or email.

Input Parameters

commitment_idstringrequiredID of the commitment to nudge about

channelstringwhatsapp | email | in_app (default: whatsapp)

messagestringCustom message to include

Example Response

{
  "nudge_id": "ndg_x1y2z3",
  "status": "sent",
  "commitment_id": "cmt_a1b2c3",
  "channel": "whatsapp",
  "sent_at": "2026-02-18T10:30:00+05:30"
}

list_workspaces

List workspaces the authenticated agent has access to.

Example Response

{
  "workspaces": [
    {
      "id": "ws_abc123",
      "name": "Mehta Textiles",
      "channels": ["whatsapp", "email"],
      "commitment_count": 47
    }
  ]
}

Agent Use Cases

Sales Operations Agent

Query at-risk commitments daily and nudge reps before promises slip.

Customer Success Agent

Monitor promise fulfillment across WhatsApp and email as a churn signal.

Supply Chain Agent

Track vendor delivery commitments from WhatsApp groups and cross-reference with ERP.

Project Management Agent

Reconcile conversational promises with task tracking systems like Jira or Linear.

← API Reference A2A Protocol →