Getting Started

Install CortexFlow-AI and have your first agent online in under a minute.

Requirements

• Python 3.12+
• Optional: Redis and Qdrant — power short-term and semantic memory; CortexFlow-AI runs with reduced memory functionality without them
• Optional: a local Ollama install — for fully offline, zero-external-API-calls privacy mode

Install

Published on PyPI — no clone required:

pip install cortexflow-ai

Or from source, for development:

git clone https://github.com/TheAmitChandra/CortexFlow-AI.git
cd CortexFlow-AI
pip install -e .

Or with Docker, no Python setup at all:

docker pull ghcr.io/theamitchandra/cortexflow-ai:latest
docker run -d -p 7432:7432 -v cortexflow-data:/root/.cortexflow \
  ghcr.io/theamitchandra/cortexflow-ai:latest

Guided setup

cortex init

Walks you through your agent's name, a primary model + API key, your first channel, and an optional voice test. Writes the result to ~/.cortexflow/config.toml. See Configuration for every available key.

Start the gateway

As a background daemon:

cortex start --background

Or in the foreground, to watch logs as messages come in:

cortex start

Talk to your agent

Directly from the terminal:

cortex chat

Or message it on whichever channel you configured in cortex init — Telegram, Discord, Slack, and 11 others are supported. See Channels.

Check on it

cortex status

Shows your agent name, primary model, gateway bind address, enabled channels, and long-term memory row count at a glance.

Every cortex command is documented on the CLI Reference page.