tbls-ask-agent-slack

This is a slack-bot version of tbls-ask.

You can call tbls-ask from slack by mentioning the bot.

Usage

To use the tbls-ask-agent-slack bot, mention it in a Slack channel where it's been added. The bot can answer questions about your database schema using natural language.

Example Query

You can ask questions like:

"@tbls-ask name of the users who has the most stars"

The bot will respond with an SQL query that answers your question. For the example above, it would provide a query to find the user with the most stars on their comments.

How It Works

1. The bot receives your question through a Slack mention.
2. It interprets your natural language query.
3. It generates an appropriate SQL query to answer your question.
4. The bot returns the SQL query in the Slack thread.

Limitations

• The bot provides SQL queries but does not execute them directly on your database.
• You need to run the provided SQL query on your own MySQL database to get the actual results.
• The database document must be prepared using tbls.

How to set up

Environment Variables

Required

• SLACK_APP_TOKEN: App token for Slack
• SLACK_OAUTH_TOKEN: OAuth token for Slack
• MODEL_NAME: Model name for LLM (default: gpt-4o)

API keys

By default, we use OpenAI models. You need to set OPENAI_API_KEY.

• OPENAI_API_KEY: API key for OpenAI

If you want to use Gemini models, you need to set GEMINI_API_KEY.

• GEMINI_API_KEY: API key for Gemini

Optional

• GITHUB_TOKEN: Token for GitHub API (optional)
• CUSTOM_INSTRUCTION: Custom instruciton for LLM (optional)
• DEBUG_MODE: When set to "true", outputs prompt contents to logs (optional)

Slack-app settings

Please create an app using manifests.yml and install it to your workspace.

Prepare schema

Copy schemas/config.yml.sample to schemas/config.yml and edit it.

$ cp schemas/config.yml.sample schemas/config.yml

Run Server

make server

This app uses socket mode for slack, so you don't need to expose the server to the internet. That means you don't need to set SLACK_SIGNING_SECRET.

Deploy to k8s

Build docker image locally

make build-image

Choose which API to use

cp manifests/deployment-openai.yml manifests/deployment.yml

Apply manifests

make all