Web Sweeper: Building Provenance-Protected Libraries for Future AI Training

Web Sweeper is an open-source library builder created by Christian Cassarly through Jesus New OS for future AI training models and research.

It is designed to move complete, policy-approved works through a protected pipeline:

1. Source discovery
2. Policy screening
3. Acquisition
4. Protected staging
5. Publication
6. Live verification

The system preserves provenance, resumability, exact duplicate protection, serialized production writes, and permanent verification receipts. Its goal is not simply to download files. Its goal is to build trustworthy, auditable digital libraries that future AI systems and researchers can use responsibly.

Why it matters

Future AI training requires more than volume. It requires clear sources, reliable processing, recoverable work, and evidence showing exactly what became part of a library. Web Sweeper brings those safeguards into one repeatable workflow.

Open-source project

Explore the source code and current documentation on GitHub:

github.com/jesusnewapp/web-sweeper

Web Sweeper is under active development. Capabilities and library totals should be evaluated from the latest public documentation and authoritative verification receipts.