About
About FoldEngine
FoldEngine exists so AI-assisted engineering can move forward without transferring release authority to the machine.
FoldEngine governs whether AI-assisted work is allowed to become real. Most tools optimize for task completion. FoldEngine optimizes for governed consequence: clarify the objective, bound the work, preserve evidence and uncertainty, wait for operator decisions, and record accepted project state.
Consequence-governance doctrine
Important AI-assisted work should be reviewable before it becomes a real change, claim, handoff, or delivery decision. Confidence is not proof. Speed is not readiness. Authority remains with the operator.
FoldEngine does not mutate source autonomously. Source changes occur only through explicit admission, authorization, bounded application, validation, and a source-state receipt. Commit, push, merge, deploy, and release remain outside autonomous authority.
Lifecycle: Research → Plan → Execute → Review → Ship. That is the public operating model — not a catalogue of internal components.
Current public products
Three independent public lanes share governance doctrine. None requires the others as a default next step. They are not equally mature expressions of the same product.
-
FoldEngine Offline Runtime
Primary product. A governed local engineering runtime for clarifying objectives, running bounded work, reviewing evidence, applying authorized changes, and preserving checkpoints under controlled delivery.
The Offline Workbench is the operator interface inside the Runtime — not a separate product and not a synonym for the entire Runtime.
-
Hosted Probes
Independent diagnostic lane. Operator-reviewed online inspections with evidence and findings — usable without installing Offline Runtime.
-
Client Portal
Independent delivery and continuity lane. Assigned-client access to approved summaries and commercial status — not a self-service report vault.
Explore the Offline Runtime, walk the Runtime Preview, or review Hosted Probes.
Local-first and human authority
FoldEngine Offline Runtime is designed to run inside a team’s own boundary. Hosted Probes and Client Portal remain online where delivery requires it. Neither lane is the hidden backend of the other.
FoldEngine is not a magic AI judge. It does not replace human responsibility. It does not silently execute release actions or claim safety certification.
Research and technical lineage
Earlier names and substrates remain part of FoldEngine’s technical history. They are not peer public products and do not change the buyer hierarchy above.
- Offline Workbench — operator interface / local workcell machinery inside FoldEngine Offline Runtime.
- PASR Local — reference / pilot packaging lineage and evaluator substrate.
- Local Reference Stack — assembled reference package of shared primitives for pilots and research.
- Runtime internals — Workcells, Nebula, Evidence, Geometry, and related mechanisms beneath the public lifecycle.
- Riverbed Atlas — later consequence-navigation projection built from runtime truth; not a primary public product.
Detailed architecture lives in documentation and the glossary, not in the primary product explanation.
See the Runtime journey
Walk one synthetic project through Offline Runtime, or request Evaluation Access for controlled deployment.