Microsoft Bot Framework SDK for Python

Within the Bot Framework, botbuilder-applicationinsights enables the Azure Application Insights service.

Application Insights is an extensible Application Performance Management (APM) service that allows you to monitor your live bot application. It provides analytics tools to help diagnose issues and understand user interactions with your bot.

CoreBot is a Bot Framework v4 sample designed to demonstrate several key bot development patterns:

• Language Understanding: Using LUIS to implement core AI capabilities.
• Multi-turn Conversations: Implementing complex flows using Dialogs.
• Interrupt Handling: Managing user interruptions such as Help or Cancel commands.
• Data Validation: Prompting for and validating specific information requested from the user.

ARCHIVE NOTICE

This repository is being archived and will no longer be updated or maintained.

• Support Status: The tool will no longer receive product updates and will not be supported through Azure portal service tickets. Support for the Bot Framework SDK will cease after December 31, 2025.
• Recommended Alternative: For building agents with AI services, orchestration, and knowledge, use the Microsoft 365 Agents SDK. The Agents SDK supports C#, JavaScript, and Python.
• Migration: If you have an existing bot, you can migrate to the Agents SDK using the migration guidance.

Install the botbuilder-schema package using pip. This package contains the serialized data sent across the wire between the user and the bot when using the Bot Framework.

pip install botbuilder-schema

Use the luis import application command to upload a local LUIS JSON model to the LUIS service.

Required parameters:

• --region: The authoring region (e.g., westus, westeurope, australiaeast).
• --authoringKey: Your LUIS Authoring Key.
• --appName: The name of your application.
• --in: Path to the local JSON file.

Important: The command returns a JSON object containing an id and activeVersion. You must save these values to use in subsequent training and publishing steps.

luis import application --region "LuisAppAuthoringRegion" --authoringKey "LuisAuthoringKey" --appName "FlightBooking" --in "./cognitiveModels/FlightBooking.json"

To consume release candidate (rc) builds of the BotBuilder SDK, you must point pip to the TestPyPI index. Use the --index-url flag to specify the TestPyPI simple index URL. Replace botbuilder-core with the specific library name you wish to install.

$ pip install --index-url https://test.pypi.org/simple/ botbuilder-core

To run the CoreBot sample locally, follow these steps:

1. Install dependencies: Run pip install -r requirements.txt to install all necessary Python packages.
2. Configure LUIS: Open config.py and update the following keys with the credentials retrieved from your LUIS portal:
   • LuisAppId
   • LuisAPIKey
   • LuisAPIHostName
3. Start the bot: Execute python app.py to launch the application.

pip install -r requirements.txt
# Update config.py with LuisAppId, LuisAPIKey, and LuisAPIHostName
python app.py

To deploy a bot using these templates, you must deploy the BotApp before the AzureBot. The deployment is performed using the Azure CLI az deployment sub create command.

Deployment Workflow:

1. Authenticate with Azure using az login.
2. Execute the deployment command providing the template file, the target region, and a parameters file.

Command Syntax:

az login
az deployment sub create --template-file <template-file> --location <bot-region> --parameters @<parameters-file>

The botbuilder-testing SDK requires the following environment configuration:

• Python: version 3.7.0 or higher.