DevHelper is a collection of AI-powered agents designed to help developers research, compare, and analyze developer tools, libraries, and platforms. The repository contains two main agents:
- advanced-agent: An advanced research agent that leverages LLMs, Firecrawl, and LangChain to extract, analyze, and recommend developer tools based on user queries.
- simple-agent: A lightweight agent that uses Firecrawl and LangChain tools via MCP for quick web scraping and information extraction.
advanced-agent/
main.py
pyproject.toml
README.md
uv.lock
src/
__init__.py
firecrawl.py
models.py
prompts.py
workflow.py
__pycache__/
simple-agent/
main.py
pyproject.toml
README.md
uv.lock
README.md
- Python 3.13+
- API keys for Firecrawl and OpenAI (add them to
.envfiles in each agent directory)
-
Clone the repository:
git clone https://github.com/yourusername/DevHelper.git cd DevHelper -
Install dependencies for each agent using uv:
cd advanced-agent uv pip install -r pyproject.toml cd ../simple-agent uv pip install -r pyproject.toml
-
Add your API keys to the
.envfiles in bothadvanced-agentandsimple-agentdirectories.
The advanced agent provides in-depth research and recommendations for developer tools.
cd advanced-agent
uv run main.py- Enter your developer query (e.g., "best Postgres alternatives for SaaS").
- The agent will extract relevant tools, analyze them, and provide concise recommendations.
The simple agent offers quick scraping and tool usage via MCP.
cd simple-agent
uv run main.py- Interact with the agent using natural language queries.
- The agent uses Firecrawl and other tools to fetch and summarize information.
- Automated extraction and comparison of developer tools
- Uses LLMs (OpenAI GPT-4) for analysis and recommendations
- Integrates with Firecrawl for web scraping and data extraction
- Modular and extensible workflow
MIT License
Thanks to the video tutorial created by TechWithTim