Control
Autonomous action needs explicit authority, bounded capabilities, and a clear human decision point.
About / Greyforge Labs
A small lab for serious software.
Greyforge is an independent software R&D lab. It builds unusual infrastructure for autonomous systems, tests the systems directly, and publishes enough evidence for the work to be inspected.
Thesis
Autonomy needs infrastructure, not just intelligence.
Inspectability
A system should expose enough structure and state to understand what it is about to change.
Verification
Claims and changes become credible when they produce repeatable, inspectable evidence.
History
Software needs durable context, retained provenance, and storage that can survive long operating horizons.
- 01Build
- 02Measure
- 03Test
- 04Publish
What we build
Flagship 01
Sley
Structural programming for compiler-mediated, human-reviewed software change.
Flagship 02
ZJX
Lossless structured-data archives with verifier-checked evidence.
Workshop
OpenForge
Focused open-source tools released with repositories, examples, and technical context.
Evidence
Research
Benchmarks, architecture studies, experiments, and field reports from the build process.
Independence
The lab model is deliberate.
Greyforge is small by design. It can pursue technically narrow ideas, reject false urgency, and let measured evidence determine which systems deserve more attention.
No invented scale. No fictional client roster. No product receives flagship status because it exists. Homepage attention is reserved for work with a clear problem, measurable capability, strategic relevance, and a path to public inspection.
Work with Greyforge