Slack Agent → Open WebUI Bridge

A Python bridge that connects Slack DMs/threads to an Open WebUI instance. Users register their Open WebUI API key via /enroll, then chat with AI models through Slack. Single Flask app, HTTP webhook mode (no sockets).

Architecture

Slack user ──DM/thread──► Single Flask app (Bolt HTTP mode)
                                 │
                                 ├─ /slack/events  → event webhook
                                 ├─ /slack/enroll  → slash command
                                 │
                                 ├─ message handler
                                 │    1. setStatus("processing")
                                 │    2. look up user API key (Postgres)
                                 │    3. create/lookup OWUI chat
                                 │    4. trigger completion
                                 │    5. post reply to Slack
                                 │    6. setStatus("active")
                                 │
                                 └─ assistant_thread_started → welcome message

One webhook URL handles everything.

Requirements

• Python 3.10+
• PostgreSQL 14+
• Running Open WebUI instance
• Slack workspace (Developer Program sandbox works)

Quick Start

1. Create a Slack app

1. Go to api.slack.com/apps → Create New App
2. Do not enable Socket Mode (we use HTTP webhook mode)
3. Under Bot Token Scopes, add:
   • chat:write
   • chat:write.public
   • im:read
   • mpim:read
   • assistant:write
4. Under Event Subscriptions, enable it, set Request URL to:

   https://<your-domain>/slack/events
   

   Subscribe to bot events:
   • message.im
   • assistant_thread_started
5. Install the app to your workspace → copy the Bot Token (xoxb-...)
6. Under Slash Commands, create one:
   • Command: /enroll
   • Request URL: https://<your-domain>/slack/events (same URL!)
   • Short Description: Register your Open WebUI API key
   • Usage Hint: <api_key>

Bolt routes both events and slash commands through the same webhook endpoint.

2. Set up the database

createdb openwebui_slack
psql openwebui_slack -f migrations/001_initial.sql

Or run the app directly — SQLAlchemy creates tables automatically.

3. Configure environment

cp .env.example .env
# Edit .env with your credentials

4. Install and run

pip install -r requirements.txt
python run.py

5. Test

In Slack:

/enroll your-openwebui-api-key

Then DM the bot with a message — it will respond via Open WebUI.

Project Structure

app/
  config.py          — Config from environment
  db.py              — SQLAlchemy engine & session
  models.py          — DB models (UserAPIKey, ThreadChatMapping)
  handlers/
    message.py       — message.im handler (core loop)
    thread.py        — assistant_thread_started welcome handler
  owui/
    client.py        — Open WebUI REST API client
    __init__.py      — exports OpenWebUIClient
migrations/
  001_initial.sql    — Initial Postgres migration
run.py               — Entry point (Bolt HTTP mode + Flask)
.env.example         — Environment template
requirements.txt     — Python dependencies

Phases

Phase	Description	Status
0	Prerequisites	✓ Setup
1	Open WebUI client library	✓ Done
2	Slack agent skeleton	✓ Done
3	Wire the loop together	✓ Done
4	Multi-turn follow-ups	Next
5	UX polish	Future
6	Hardening	Future
7	Test & ship	Future