Slack QA (Question answering) bot

This is a fork of https://github.com/seratch/ChatGPT-in-Slack with additional langchain and integration with https://github.com/go-skynet/LocalAI to run completely locally, without need of any OpenAI account.

The bot allows to ask queries and uses langchain to build a vector database from a website documentation, or a list of Github projects.

How It Works

You have to start a new thread by mentioning the bot: @ What's ...?

Note: It does not reply to mentions. That was done in purpose to avoid to send big context back to the LLM, which makes it slow when running entirely on CPU.

Running the App on Your Local Machine

To run this app on your local machine, you only need to follow these simple steps:

• Create a new Slack app using the manifest-dev.yml file
• Install the app into your Slack workspace
• Retrieve your OpenAI API key at https://platform.openai.com/account/api-keys
• Start the app

# Create an app-level token with connections:write scope
export SLACK_APP_TOKEN=xapp-1-...
# Install the app into your workspace to grab this token
export SLACK_BOT_TOKEN=xoxb-...
# Visit https://platform.openai.com/account/api-keys for this token
export OPENAI_API_KEY=sk-...

# Optional: gpt-3.5-turbo and gpt-4 are currently supported (default: gpt-3.5-turbo)
export OPENAI_MODEL=gpt-4
# Optional: You can adjust the timeout seconds for OpenAI calls (default: 30)
export OPENAI_TIMEOUT_SECONDS=60


export MEMORY_DIR=/tmp/memory_dir

export OPENAI_API_BASE=http://localhost:8080/v1

export EMBEDDINGS_MODEL_NAME=all-MiniLM-L6-v2

## Repository and sitemap to index in the vector database on start
export SITEMAP="https://kairos.io/sitemap.xml"
export REPOSITORIES="foo,bar"
export foo_CLONE_URL="http://github.com.."
export bar_CLONE_URL="..."
export foo_BRANCH="master"
export GITHUB_PERSONAL_ACCESS_TOKEN=""
export ISSUE_REPOSITORIES="go-skynet/LocalAI,foo/bar,..."

# Optional: When the string is "true", this app translates ChatGPT prompts into a user's preferred language (default: true)
export USE_SLACK_LANGUAGE=true
# Optional: Adjust the app's logging level (default: DEBUG)
export SLACK_APP_LOG_LEVEL=INFO
# Optional: When the string is "true", translate between OpenAI markdown and Slack mrkdwn format (default: false)
export TRANSLATE_MARKDOWN=true

python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
python main.py

Running the App for Company Workspaces

Confidentiality of information is top priority for businesses.

This app is open-sourced! so please feel free to fork it and deploy the app onto the infrastructure that you manage. After going through the above local development process, you can deploy the app using Dockerfile, which is placed at the root directory of this project.

The Dockerfile is designed to establish a WebSocket connection with Slack via Socket Mode. This means that there's no need to provide a public URL for communication with Slack.

Contributions

You're always welcome to contribute! 🙌 When you make changes to the code in this project, please keep these points in mind:

• When making changes to the app, please avoid anything that could cause breaking behavior. If such changes are absolutely necessary due to critical reasons, like security issues, please start a discussion in GitHub Issues before making significant alterations.
• When you have the chance, please write some unit tests. Especially when you touch internals.py and add/edit the code that do not call any web APIs, writing tests should be relatively easy.
• Before committing your changes, be sure to run ./validate.sh. The script runs black (code formatter), flake8 and pytype (static code analyzers).

The License

The MIT License