AnswerRocket Python SDK ( answerrocket-client ) Integration Guide

This guide walks third-party developers through integrating with AnswerRocket using its official Python SDK: answerrocket-client. The SDK enables secure, programmatic access to AnswerRocket’s APIs, data sources, and AI-powered assistants.

📦 Installation (via pip)

Install the SDK using Python’s package manager:

pip install answerrocket-client

It’s recommended to use a virtual environment to isolate dependencies.

🔐 Authentication & Setup (Token-Based)

To securely connect to AnswerRocket using an API token:

from answer_rocket import AnswerRocketClient

arc = AnswerRocketClient(
    url='https://your-answerrocket-instance.com',
    token='<your_api_token>'
)

# Confirm the credentials work
arc.can_connect()

Replace https://your-answerrocket-instance.com with your actual AnswerRocket environment URL.
Replace <your_api_token> with your generated API token.

🔑 How to Generate Your API Token

To generate an API token:

Click the User Icon in the bottom-left corner of the AnswerRocket interface.
In the profile panel, find the Client API Key section.
Click Regenerate to generate a new token or Revoke to disable an existing one.
Copy and use the token in your code.

🧠 SDK Capabilities

The answerrocket-client SDK includes:

1. Configuration Management

Access your current assistant (copilot) and its registered skills
Manage chat questions and link them to skills
Load and interact with artifacts (like layout templates and resource files)

2. Data Access

Execute SQL queries against connected databases
Retrieve structured results as pandas DataFrames

3. Skill Development

Access the runtime context for skills
Store and retrieve payloads for evaluations, visualizations, or debugging

4. AI and Language Model Services

Generate chat completions using AnswerRocket's LLMs
Stream real-time model responses

5. GraphQL API Integration

Interact with the platform’s GraphQL schema using the built-in client
Supports introspection and typed responses

For full documentation and additional examples, refer to our GitHub wiki:
📖 AnswerRocket Python SDK Wiki

💾 Example: Save a Custom Payload in a Skill

import pandas as pd
from answer_rocket.skill_platform import load_skill_platform

sp = load_skill_platform()
save_area = sp.ctx.get_payload_save_area()

save_area["validation_df"] = pd.DataFrame({
    "user_question": ["barilla sales in 2022"],
    "data": ["{}"]
})

This enables storing structured content like validation data for downstream processing or UI rendering.

For further guidance, visit the AnswerRocket SDK Wiki for deep dives into advanced usage, code samples, and architectural concepts.