orqAgent

Introduction

orqAgent is a no-code platform that lets you build custom AI agents powered by models like Claude, GPT-4, and Gemini. Each agent has its own personality, knowledge base, guardrails, and REST API endpoint — ready to embed in any product in minutes.

Quick Start

From sign-up to your first API call in under 5 minutes.

curl -X POST https://your-domain.com/api/agents/{agent_id}/run \
  -H "Content-Type: application/json" \
  -H "X-API-Key: orq_your_api_key" \
  -d '{"user_input": "Hello!"}'

Creating an Agent

An agent is the core building block. It bundles a model, a personality, and optional capabilities (knowledge base, guardrails, rate limits) into a single deployable unit.

Basic settings

Persona fields

Persona fields let you shape how the agent behaves without cluttering the system prompt.

Advanced options

Playground

The Playground is a fully-featured chat interface wired directly to your agent — same model, same system prompt, same knowledge base as the live API. Use it to iterate on prompts before going to production.

Knowledge Base

Upload documents and your agent automatically retrieves the most relevant chunks for each user question. This is done via retrieval-augmented generation (RAG) — no manual prompt engineering required.

Supported formats

How it works

API Keys

API keys let your backend or external tools call any of your agents over HTTP without going through the dashboard UI.

Creating a key

Key properties

REST API Reference

Every agent exposes a simple HTTP endpoint. Authenticate with your API key in theX-API-Key header.

Authentication

X-API-Key: orq_your_api_key_here

Request body

Response body

Example — cURL

curl -X POST https://your-domain.com/api/agents/YOUR_AGENT_ID/run \
  -H "Content-Type: application/json" \
  -H "X-API-Key: orq_your_api_key" \
  -d '{
    "user_input": "What is your return policy?",
    "session_id": "user-abc-session-1"
  }'

Example — JavaScript (fetch)

const response = await fetch(
  `https://your-domain.com/api/agents/${AGENT_ID}/run`,
  {
    method: "POST",
    headers: {
      "Content-Type": "application/json",
      "X-API-Key": process.env.ORQ_API_KEY,
    },
    body: JSON.stringify({
      user_input: userMessage,
      session_id: sessionId,   // keep across turns for memory
    }),
  }
);

const data = await response.json();
console.log(data.agent_response);

Example — Python

import httpx, os

BASE_URL = "https://your-domain.com"
AGENT_ID = "your-agent-id"
API_KEY  = os.environ["ORQ_API_KEY"]

def chat(user_input: str, session_id: str) -> str:
    r = httpx.post(
        f"{BASE_URL}/api/agents/{AGENT_ID}/run",
        headers={"X-API-Key": API_KEY},
        json={"user_input": user_input, "session_id": session_id},
        timeout=30,
    )
    r.raise_for_status()
    return r.json()["agent_response"]

reply = chat("Summarise your pricing plans.", "session-42")
print(reply)

Public agent endpoint (no auth)

If you've toggled Public on an agent, it can be called without an API key. Usage is still billed to your account.

# No API key needed — works from any browser or server
curl -X POST https://your-domain.com/api/public/agents/YOUR_AGENT_ID/chat \
  -H "Content-Type: application/json" \
  -d '{"user_input": "Hello!"}'

Error codes

Guardrails

Guardrails are automatic input filters that block toxic, harmful, or off-topic messages before they reach the LLM — protecting your quota and your users.

What gets blocked

When a guardrail fires

The API returns HTTP 400 with a clear error message. The run is still logged with blocked_by_guardrail: true so you can track block rates in the Analytics tab.

{
  "error": "Input blocked by content guardrail",
  "code":  "GUARDRAIL_BLOCKED"
}

Plans & Limits