Quick Start Guide

After signing up, go to Settings > API in your dashboard to find your API key.

API_KEY = "vd_your_api_key_here"

Register your AI agent so Vector Decisions knows about it.

import requests

response = requests.post(
    "https://vectordecisions.com/api/v1/agents",
    headers={"Authorization": f"Bearer {API_KEY}"},
    json={
        "name": "My Support Bot",
        "type": "SUPPORT",
        "purpose": "Handles customer support inquiries",
        "vendor": "OpenAI",
        "model": "gpt-4",
        "riskLevel": "LIMITED"
    }
)

agent_id = response.json()["id"]
print(f"Agent registered: {agent_id}")

As your agent performs actions, log them to Vector Decisions.

requests.post(
    "https://vectordecisions.com/api/v1/events",
    headers={
        "Authorization": f"Bearer {API_KEY}",
        "X-Agent-ID": agent_id
    },
    json={
        "eventType": "ACTION",
        "action": "Responded to customer inquiry about billing",
        "category": "COMMUNICATE",
        "duration": 1200,
        "success": True
    }
)

Your agent should periodically check its status. If the dashboard operator pauses or stops the agent, the status endpoint will reflect that.

import time

while True:
    # Check status every 30 seconds
    status = requests.get(
        f"https://vectordecisions.com/api/v1/agents/{agent_id}/status",
        headers={"Authorization": f"Bearer {API_KEY}"}
    ).json()

    if not status["shouldContinue"]:
        print(f"Agent status: {status['status']} - halting operations")
        break

    # ... your agent does its work here ...

    time.sleep(30)