🛠️ Installation Guide

This guide walks you through setting up and running the co_ai framework locally.

---

📦 Requirements

• Python 3.8+
• PostgreSQL with pgvector extension
• Ollama (for local LLM inference)
• Poetry OR standard pip + venv
• Optional: Docker (for running PostgreSQL locally)

---

🔧 1. Clone the Repository

git clone https://github.com/ernanhughes/co-ai.git
cd co-ai
````

---

## 🐍 2. Create a Virtual Environment

**Using `venv`:**

```bash
python3 -m venv .venv
source .venv/bin/activate   # On Windows: .venv\Scripts\activate

Or with poetry:

poetry install
poetry shell

---

📚 3. Install Dependencies

pip install -r requirements.txt

OR using pyproject.toml:

pip install .

---

🧠 4. Set Up PostgreSQL + pgvector

Option A: Using Docker

docker run --name coai-db -p 5432:5432 \
  -e POSTGRES_PASSWORD=postgres \
  -e POSTGRES_USER=postgres \
  -e POSTGRES_DB=coai \
  ankane/pgvector

Option B: Local Installation

1. Install PostgreSQL

2. Install pgvector:

bash CREATE EXTENSION vector;

1. Create the database:

bash createdb coai

1. Run the schema:

bash psql -U postgres -d coai -f schema.sql

---

🤖 5. Install & Run Ollama

ollama run qwen:latest

Or for smaller models:

ollama run llama2

Make sure Ollama is running on http://localhost:11434.

---

⚙️ 6. Run the App

python co_ai/main.py goal="The USA is on the verge of defaulting on its debt"

Or with CLI args:

python co_ai/main.py --config-name=config goal="My research goal here"

---

📝 Notes

• Logs are stored in logs/ as structured JSONL.
• Prompts are saved in prompts/ and tracked in the database.
• You can inspect all configuration using Hydra or customize each agent via config/.

---

✅ You’re Ready!

You now have a full pipeline for running research-style hypothesis generation, evaluation, and prompt tuning.