title	emoji	colorFrom	colorTo	sdk	sdk_version	app_file	pinned	hf_oauth	hf_oauth_expiration_minutes
Template Final Assignment	🕵🏻‍♂️	indigo	indigo	gradio	5.25.2	app.py	false	true	480

HuggingFace Agents Course Final Project

Final Sumbission for the Hugging Face Agents Course

Score Achieved: 90%

Requirements

Install Python 3 and pip

python3 -m venv ./venv
source ./venv/bin/activate
pip install -r requirements.txt

Install Stockfish to solve the chess problem

sudo apt install stockfish

Install Firefox to enable Selenium for the Research Agent

Firefox for Linux Installation Guide

Optional

Langfuse is the desired application of choice for observability. This project can be run without langfuse but its easy to set up langfuse locally using docker

API Keys

API Keys needed to run the project

1. OpenRouter
2. OpenAI
3. Brave Search

Optional

4. Langfuse Public & Secret Keys

.env File

The .env file should look like this

SPACE_ID=
OPENROUTER_KEY=
BRAVE_KEY=
OPENAI_API_KEY=

# Optional Requirements
LANGFUSE_HOST=
LANGFUSE_PUBLIC_KEY=
LANGFUSE_SECRET_KEY=

Run Tests

The Agent can be tested on a group of selected tasks out of the 20 or all tasks. To do this edit the test_questions list in the test_workflow.py file. Then run the following command

deepeval test run test_workflow.py

Run the Agent

To start the app run the following command:

python3 app.py

Navigate to http://127.0.0.1:7860

The Course

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference