Too Long; Didn't Scroll (TL;DS)

A Slack bot that helps you to summarize long messages and threads.

Features

• Summarize messages and threads
  • Running on a top-level message will summarize the message along with its replies, if any exist
  • Running on a reply will summarize the reply
• Summarize messages and threads with a single click (well, two clicks)
  • Use the Generate Summary message action to summarize
• Customizable visibility
  • When using the /tlds command, you can choose whether the summary is visible to everyone or is ephemeral and only visible to you
  • When using the Generate summary message action, the summary is ephemeral and only visible to you
  • When using the Generate public summary message action, the summary is visible to everyone
• Customizable LLM model and provider
  • Supports Ollama for local/on-premises summarization
  • Supports OpenAI-compatible APIs for cloud-based summarization
    • Examples include OpenAI and OpenRouter
• A help command (/tlds-help) to show the available commands and options

Setup

1. Create a new Slack app at api.slack.com/apps
2. Use the manifest.yml file to create it
3. Install the app to your workspace
4. Clone the repository

5. Copy the example.env file to .env and fill in the values
   • Alternatively, you can use .secrets.toml and settings.toml
   • For more information, check the example.*.toml files

Running the server

Docker

Run the following command to build and run the Docker container:

docker-compose up

If you want to run the container in the background, use the -d flag:

docker-compose up -d

If you have Ollama running on the host, the host can be accessed by the container by using host.docker.internal as the host name. For example, if Ollama is running on port 11434 (the default), setting the base_url to http://host.docker.internal:11434 will allow the container to access Ollama. The docker-compose.yml file also has a commented-out section for Ollama that can be used. However, setting it up on the host or by itself can offer more flexibility. To use the Ollama section in the docker-compose.yml file, uncomment the section. The base_url should be http://ollama:11434 in this case.

Poetry

If you want to run the server without Docker, you can use Poetry:

poetry install
poetry run python -m tlds

Usage

1. Add the app to channel(s) where you want it to be able to summarize messages from
2. Use the /tlds command or the Generate Summary/Generate public summary message actions to summarize messages and threads