# Building a Slackbot

Set up a self-hosted Warp Slackbot that answers repo questions and opens PRs directly from Slack using Docker and GitHub integration.

This guide shows you how to spin up a [Warp Slackbo**t**](https://github.com/warpdotdev/warp-slackbot-public) you can mention in Slack to answer questions about your repositories and even open pull requests. It runs securely in Docker and connects directly to your Slack workspace.

The setup takes just a few steps — clone, configure, and run — and requires minimal setup beyond providing your own Slack, GitHub, and Warp credentials.

## Why this is useful

-   Run a self-hosted Slack bot that connects your team’s repos and Warp AI agents.
-   Provide your team with a coding assistant that can answer repo questions or help with PRs directly in Slack.

## Quickstart Setup

1.  #### Clone the repository
    
    Clone the public repo and navigate into it:
    
    ```
    git clone https://github.com/warpdotdev/warp-slackbot-public.gitcd warp-slackbot-public
    ```
    
    This repository includes all required configuration templates, the Docker setup, and the app manifest you’ll use to create your Slack bot.
    
2.  #### Configure environment variables
    
    Copy the provided example environment file and fill in your credentials:
    
    ```
    cp .env.example .env
    ```
    
    Then open `.env` and fill in the following values:
    
    -   `SLACK_BOT_TOKEN` – Your bot token from Slack, which starts with `xoxb-`. You can find this under your Slack app’s **OAuth & Permissions** page.
    -   `SLACK_APP_TOKEN` – Your app-level token from Slack, which starts with `xapp-`. Create this in your Slack app’s **Basic Information → App-Level Tokens**, and ensure it has the `connections:write` scope (required for Socket Mode).
    -   `GITHUB_PAT` – Your GitHub Personal Access Token (PAT) with `repo` access, so the bot can read and clone your repositories.
    -   `WARP_API_KEY` – Your Warp API key that allows the bot to connect to Warp’s agentic environment and execute prompts.
    
    You only need to create these once — after saving your `.env` file, Docker will automatically read them when you run the bot.
    
3.  #### Configure repositories
    
    Copy the provided repository configuration template:
    
    ```
    cp repos.yaml.template repos.yaml
    ```
    
    Then open `repos.yaml` and list the repositories you want the bot to monitor, for example:
    
    ```
    repositories:  - url: "myorg/backend"    branch: "main"  - url: "myorg/frontend"    branch: "develop"
    ```
    
    This file tells the bot which repositories to clone locally and keep up to date. It’s how the bot knows where to pull context from when you ask it repo-related questions.
    
    What’s happening behind the scenes:
    
    -   The bot will use your GitHub PAT to clone each listed repo.
    -   It stores the repositories in a persistent Docker volume so you don’t have to re-clone each time.
    -   The bot indexes those repos for context so it can respond intelligently when mentioned in Slack.
4.  #### Understand the Slack App Manifest
    
    The file `slack_app_manifest.json` defines everything needed to create your Slack app quickly — including permissions, event subscriptions, and bot scopes. When creating your Slack app, you can paste the contents of this file directly into the **Slack App → Create from Manifest** interface. It configures the bot to:
    
    -   Listen for mentions and thread replies (`app_mention` events)
    -   Run via Socket Mode (secure WebSocket connection)
    -   Send and receive messages in channels where it’s invited
5.  #### Run the bot in Docker
    
    Once your `.env` and `repos.yaml` files are ready, start the bot:
    
    ```
    docker-compose up --build
    ```
    
    This builds the container, authenticates with Slack and GitHub, and starts listening for messages. It will log all activity in your terminal so you can confirm it’s running.
    
6.  #### Test in Slack
    
    -   Invite your bot to a channel in Slack.
    -   Mention it directly or in a thread to trigger it — for example:
    
    ```
    @Warp analyze the recent changes in the main branch@Warp help me review this PR
    ```
    
    -   The bot will pull context from your configured repositories and reply with AI-assisted insights using Warp’s agentic platform.

## What happens on startup

-   Reads `.env` and `repos.yaml`.
-   Authenticates to Slack (Socket Mode and Web API).
-   Authenticates to GitHub and clones the listed repos.
-   Starts listening for `app_mention` events and threaded messages.
-   Routes context and commands to Warp’s AI agent backend.

You can stop the bot anytime with `Ctrl + C` or run it persistently with:

```
docker-compose up -d
```