This project uses Ollama to run a local AI model on your computer.

Install Ollama from:

https://ollama.com

On Linux, you can install Ollama with:

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

Check that Ollama installed:

ollama --version

If Ollama is not running, start it with:

sudo systemctl start ollama

Optional status check:

sudo systemctl status ollama

This template uses Qwen3 4B by default:

ollama pull qwen3:4b

You can test the model directly with:

ollama run qwen3:4b

Then try a simple prompt:

Say hello in one sentence and confirm the local model is working.

To exit the Ollama chat, type:

/bye

Clone the project from GitHub:

git clone https://github.com/YOUR-USERNAME/local-ai-work-agent-template.git

Move into the project folder:

cd local-ai-work-agent-template

Replace YOUR-USERNAME with your actual GitHub username or organization name.

Create the virtual environment:

python3 -m venv .venv

Activate it on Linux or macOS:

source .venv/bin/activate

Activate it on Windows PowerShell:

.venv\Scripts\Activate.ps1

When the virtual environment is active, you should see something like this at the beginning of your terminal line:

(.venv)

Install the required packages:

pip install -r requirements.txt

For this first version, the project uses:

requests
python-dotenv

requests lets Python send a request to Ollama.

python-dotenv lets Python read local settings from the .env file.

Copy the example environment file:

cp .env.example .env

Your .env file should include:

OLLAMA_MODEL=qwen3:4b
OLLAMA_BASE_URL=http://localhost:11434
AGENT_NAME=Local Work Agent

ENABLE_THINKING=false
ENABLE_FILE_ACCESS=true
ENABLE_CALENDAR_ACCESS=false
ENABLE_WEATHER_ACCESS=false

DATA_INBOX_PATH=data/inbox
DATA_PROCESSED_PATH=data/processed
DATA_OUTPUTS_PATH=data/outputs

ALLOW_FULL_COMPUTER_ACCESS=false
SAVE_OUTPUTS=true

The .env.example file is safe to commit to GitHub.

The real .env file should stay on your computer and should not be pushed to GitHub.

From the project root, with your virtual environment activated, run:

python3 -m app.main

Expected output:

Local Work Agent is starting...
Using local model: qwen3:4b
Thinking enabled: False

Agent response:
Local model connected.

This proves the basic local path is working:

Python project
→ Ollama running locally
→ Qwen model
→ response back to Python

On some Linux systems, use python3 instead of python.

python3 -m venv .venv

If you see an error about venv not being available, install it:

sudo apt-get update
sudo apt-get install python3-venv

Then try again:

python3 -m venv .venv

If you see:

ModuleNotFoundError: No module named 'requests'

Make sure your virtual environment is activated:

source .venv/bin/activate

Then install dependencies:

pip install -r requirements.txt

If you see:

ERROR: This version requires zstd for extraction.

Install zstd first:

sudo apt-get update
sudo apt-get install zstd

Then rerun the Ollama install command:

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

Qwen3 can sometimes generate reasoning text even for simple prompts.

For this first connection test, the template uses several guardrails:

• ENABLE_THINKING=false
• /no_think
• a strict system prompt
• low randomness settings
• a response length limit
• cleanup logic for leaked thinking text

The goal of the first test is not deep reasoning.

The goal is to confirm that Python can talk to Ollama and return a clean local response.

Make sure your requirements.txt has:

requests
python-dotenv

Make sure your .gitignore includes:

.env
.venv/

data/inbox/*
data/processed/*
data/outputs/*

!data/inbox/.gitkeep
!data/processed/.gitkeep
!data/outputs/.gitkeep