Learn the accepted Agent-first behavior, system boundaries, and laboratory architecture.
Appearance
Welcome to RIGOR
RIGOR is an Agent-first laboratory operating system. A researcher defines a scientific goal, constraints, and an autonomy policy; RIGOR plans and executes routine work, verifies evidence, analyzes results, and advances the next iteration. People step in for important decisions, required physical work, or failures that automation cannot recover.
Loading the latest laboratory view…
Live system statusUpdated waiting for first update · refreshes every 15 seconds
—devices online / registered
—devices intentionally disabled
—core services healthy
—active operations and human tasks
Choose Your Path
Start with the task that brought you here. Each route points to maintained current documentation rather than historical design material.
Start the required services, investigate an end-to-end failure, and qualify a release.
Find the correct source of truth, repository rules, and contributor workflow.
Review qualified baselines, benchmark studies, and dated release records.
How the System Works
The reasoning and execution layers remain deliberately separate. Models can propose what should happen; deterministic services retain authority for policy, validation, locking, timeout, cancellation, safe-stop, verification, and evidence commitment.
IntentResearcherGoal and policy
CoordinateASCENDStudy and decisions
PlanComposerValidated RunSpec
ExecutePACE + LabFlowDeterministic work
OperateLabBridgeDevices and evidence
AnalyzePRISMCommitted results
Current Capability
- ASCEND provides the researcher interface and one
Needs yousurface. - Atomic actions and reproducible workflows share policy, audit, and evidence boundaries.
- Rehearsal supports isolated branches, comparison, promotion, and truthful 2D/3D playback.
- Restart, durable waits, result delivery, callbacks, and evidence replay have qualification coverage.
Institution identity, target PostgreSQL restoration, redundant object storage, and independently monitored off-host backup remain open. Physical devices and effects require a later accepted PRD and are outside the current release scope.
Review maintained progress →Project Team
Core developers own the platform's core design and engineering implementation. Di Zhang leads the project, product direction, and overall delivery; Hao Li provides scientific oversight and guidance on research direction.
Core DeveloperYiming Lu
Core DeveloperTingyu Lu
Core DeveloperLione Sebastian Veiga
Project LeadDi Zhang
Project AdvisorHao Li
Reading and Evidence Conventions
RequirementsAccepted product behaviorProgressImplemented and remaining workRelease recordsDated qualification evidenceCompleted designsHistorical, not a roadmapVision and visualsIntent, not implementation evidence
English is the primary language for current Web documentation. Translations are maintained as separate editions; historical sources and visual briefs do not override current requirements. This separation prevents an old plan, a visual concept, or a successful simulation from being mistaken for production or physical-laboratory readiness.