Cortex — Discord Daily Summary Bot

A production-ready Discord bot that aggregates messages from configured channels, summarizes them via local Ollama, and DMs you daily and weekly.

Features

• Daily summary — Messages from configured channels summarized and sent via DM at a scheduled time
• Weekly summary — Last 7 days of daily summaries aggregated into a weekly digest
• 7-day retention — Summaries stored in SQLite for weekly rollup; old data pruned automatically

Requirements

• Python 3.11+
• Ollama running locally with a model (e.g. ollama pull llama3.2)
• Discord bot token and server access for the channels you want to monitor

Setup

1. Create a Discord application and bot

2. Invite the bot to your servers with "Read Message History" and "Send Messages" (for DMs)

3. Install Ollama and pull a model: ollama pull llama3.2

4. Copy .env.example to .env and fill in values:

   cp .env.example .env

5. Run with Docker:

   docker-compose up -d

   Or run locally:

   python -m venv .venv
   source .venv/bin/activate  # or `.\\.venv\\Scripts\\activate` on Windows
   pip install -r requirements.txt
   python -m src.main

Configuration

Variable	Description
DISCORD_TOKEN	Bot token
DISCORD_USER_ID	Your Discord user ID (recipient of summaries)
CHANNEL_IDS	Comma-separated channel IDs to monitor
OLLAMA_HOST	Ollama API URL (use http://host.docker.internal:11434 in Docker to reach host Ollama)
OLLAMA_MODEL	Model name (e.g. llama3.2, mistral)
SUMMARY_UTC_HOUR	Hour for daily summary (0–23)
SUMMARY_UTC_MINUTE	Minute for daily summary
SUMMARY_WEEKLY_DAY	Day of week for weekly summary (0=Mon, 6=Sun)
SUMMARY_WEEKLY_UTC_HOUR	Hour for weekly summary
SUMMARY_WEEKLY_UTC_MINUTE	Minute for weekly summary

Limitations

• Bot must be in the server and have Read Message History in each monitored channel
• Ollama must be reachable from the container (e.g. host.docker.internal on Mac/Windows)
• Only the last 500 messages per channel per run are collected

Troubleshooting: "Failed to connect to Ollama" in Docker

If Cortex runs in Docker and cannot reach Ollama:

1. Ollama must listen on all interfaces — By default Ollama binds to 127.0.0.1, which is unreachable from a container. On the host, start Ollama with:

   OLLAMA_HOST=0.0.0.0 ollama serve

   Or set OLLAMA_HOST=0.0.0.0 in your shell before running ollama serve. If Ollama runs as a service, configure it to listen on 0.0.0.0.

2. Use the right OLLAMA_HOST in .env — Inside the container, localhost refers to the container, not your machine. In .env set:

   OLLAMA_HOST=http://host.docker.internal:11434
   

   (Mac/Windows Docker Desktop). On Linux, host.docker.internal is not defined by default — add to docker-compose.yml under the service:

   extra_hosts:
     - "host.docker.internal:host-gateway"

3. Verify Ollama is reachable — On the host: curl http://localhost:11434/api/tags. From inside the container:

   docker exec cortex curl -s http://host.docker.internal:11434/api/tags

   A JSON response means Ollama is reachable.

Ollama as a systemd service (Linux, runs in background, survives reboot)

If Ollama is not installed or you want a proper systemd service:

1. Install Ollama (if needed):

   curl -fsSL https://ollama.com/install.sh | sh

   This creates the ollama user and installs the systemd unit.

2. Stop any running Ollama so port 11434 is free:

   sudo systemctl stop ollama
   pkill ollama   # fallback if not managed by systemd

3. Configure Ollama to listen on all interfaces (required for Docker). Use an override file so updates do not overwrite your config:

   sudo mkdir -p /etc/systemd/system/ollama.service.d
   printf '[Service]\nEnvironment="OLLAMA_HOST=0.0.0.0"\n' | sudo tee /etc/systemd/system/ollama.service.d/override.conf

4. Reload, enable, and start:

   sudo systemctl daemon-reload
   sudo systemctl enable ollama
   sudo systemctl start ollama

5. Verify — Ollama should listen on 0.0.0.0:11434:

   ss -tlnp | grep 11434
   # Expected: 0.0.0.0:11434 (not 127.0.0.1)
   curl -s http://localhost:11434/api/tags

6. Pull a model (once):

   ollama pull llama3.2

Useful commands:

• sudo systemctl status ollama — check status
• sudo systemctl restart ollama — restart after config changes
• sudo journalctl -u ollama -f — follow logs