🔗 Universal Agent Connector

Connect any AI agent to CredEx in under 5 minutes. No SDK required. Any language. Any framework. Just a webhook.

How It Works

Webhook Contract

CredEx sends a POST to your webhook URL:

{
  "message": "What were AAPL's Q1 earnings?",
  "conversation_history": [
    { "role": "user", "content": "I'm researching tech stocks" },
    { "role": "assistant", "content": "I can help with that..." }
  ],
  "memories": [
    { "content": "User is interested in AAPL and MSFT", "context": null }
  ],
  "agent_name": "My Research Agent",
  "metadata": {
    "platform": "credexai",
    "timestamp": "2026-05-28T14:00:00.000Z"
  }
}

Your endpoint returns:

{ "response": "AAPL reported Q1 revenue of $124.3B..." }

Or just plain text. Any JSON field named response, content, message, text, output, or reply works.

Quick Start

Option A: Through the UI

1. Go to Agent Hub
2. Click "+ Add Agent"
3. Select "🔗 Connect Any Agent"
4. Enter your webhook URL and optional auth token
5. Click "🧪 Test Connection"
6. Click "Connect Agent"

Option B: Via API

# Register your external agent
curl -X POST https://credexai.live/api/agents \
  -H "Authorization: Bearer YOUR_JWT_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "My LangChain Agent",
    "type": "imported_webhook",
    "webhook_url": "https://your-server.com/agent/webhook",
    "auth_token": "optional-bearer-token"
  }'

Option C: Via MCP

If your framework supports MCP, CredEx is discoverable at:

GET https://credexai.live/.well-known/mcp.json

Use the credex_connect_agent tool to register programmatically.

Examples

from flask import Flask, request, jsonify

app = Flask(__name__)

@app.route("/webhook", methods=["POST"])
def handle():
    data = request.json
    message = data["message"]
    memories = data.get("memories", [])
    history = data.get("conversation_history", [])

    # Your agent logic here
    response = my_agent.run(message, context=memories, history=history)

    return jsonify({"response": response})

if __name__ == "__main__":
    app.run(port=8080)

const express = require('express');
const app = express();
app.use(express.json());

app.post('/webhook', async (req, res) => {
  const { message, memories, conversation_history } = req.body;

  // Your agent logic here
  const response = await myAgent.run(message, { memories, history: conversation_history });

  res.json({ response });
});

app.listen(8080);

from flask import Flask, request, jsonify
from langchain.agents import AgentExecutor
from langchain_openai import ChatOpenAI

app = Flask(__name__)
# ... set up your LangChain agent ...

@app.route("/webhook", methods=["POST"])
def handle():
    data = request.json

    # Inject CredEx memories as context
    memory_ctx = "\n".join(m["content"] for m in data.get("memories", []))
    enriched = f"Context:\n{memory_ctx}\n\nUser: {data['message']}"

    result = agent_executor.invoke({"input": enriched})
    return jsonify({"response": result["output"]})

from flask import Flask, request, jsonify
from crewai import Agent, Task, Crew

app = Flask(__name__)

@app.route("/webhook", methods=["POST"])
def handle():
    data = request.json

    agent = Agent(role="Analyst", goal="Accurate analysis", ...)
    task = Task(description=data["message"], agent=agent)
    crew = Crew(agents=[agent], tasks=[task])

    result = crew.kickoff()
    return jsonify({"response": str(result)})

Requirements

Authentication & Headers

CredEx sends these headers on every webhook call:

Testing Your Webhook

curl -X POST https://credexai.live/api/agents/test-webhook \
  -H "Authorization: Bearer YOUR_JWT_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "webhook_url": "https://your-server.com/agent/webhook",
    "auth_token": "optional"
  }'

# Response:
{
  "success": true,
  "response_time_ms": 245,
  "preview": "Hello! This is my agent responding..."
}

Discovery Endpoints

FAQ