Welcome to an experimental, open-source autonomous agent framework. This project is designed to perform zero-cost B2B data extraction while utilizing primitive "Self-Healing" and "Code-Refining" logic—stepping stones toward Artificial General Intelligence (AGI).
Most web scrapers break when UI changes or firewalls block them, requiring a human developer to fix the code. This project introduces a Self-Refinery Module. When the script encounters a failure (like a Timeout or a Block), it attempts to read its own error logs and dynamically rewrite its own Python code to overcome the obstacle.
By open-sourcing this, we invite developers and AI researchers to test, break, and evolve the architecture.
- The Autonomous Hustler (
lead_generator.py): Hunts for hyper-targeted B2B leads via search engine dorking without needing expensive API keys. - The Self-Refinery (
self_refinery.py): The immune system. Scans error logs and mutates source code (e.g., changing headers, increasing timeouts) to adapt to resistance. - Global Intelligence Sweep (
intelligence_gatherer.py): Scans hacker/tech forums to learn new trends and writes strategic reports. - Cloud Spores (
.github/workflows): A fully automated CI/CD pipeline that allows the agent to wake up every hour, scrape data, self-heal, and commit the extracted CSV files back to the repository.
This is a transparent, community-driven effort to build autonomous cognitive architectures.
- Clone the repository.
- Run the modules locally.
- Watch the agent adapt.
Disclaimer: This is an experimental framework for educational and research purposes in cognitive architectures and autonomous code mutation.