Sustainable Catalyst Scientific Workspace
Lab
Scientific modeling, data analysis, visualization, experiments,
evidence, notebooks, and reproducible computational research.
v0.48.0
Model. Visualize. Test. Record.
Use the active project as a connected scientific workspace. Start with a model or observation, inspect the evidence visually, then preserve the result as a reproducible project record.
Multidimensional response field Illustrative
Explore a browser-rendered 4D response surface, vector field, uncertainty envelope, contours, and projected tesseract before opening Graph Studio.
Response surface configuration
Uncertainty, sensitivity & ensemble diagnostics
Monte Carlo intervals, local finite-difference sensitivities, and seed-replication ensembles are modeled diagnostics. They do not establish causality, statistical significance, observed evidence, calibrated forecast skill, or scientific validity. Formal Sobol/Morris analysis remains available in the specialist Sensitivity & Global Uncertainty Analysis Studio.
Linked Scientific Views & Cross-View Selection
Selection inspector
- Source view
- No selection
- Model / mode
- —
- X
- —
- Y
- —
- W
- —
- Z / value
- —
Parameter snapshot
| Parameter | Value | Role |
|---|---|---|
| Select a compute-backed surface point to inspect parameters. | ||
Selection history
- No selections yet.
Linked selection synchronizes inspection state across the 4D response surface, parameter snapshot, and computed diagnostic views. Diagnostic selections are preserved with their own model identity and are not automatically mapped onto a different response-surface model. Graph Studio handoff transfers the selected state and provenance; it does not assert equivalence, causality, evidence, significance, calibration, or scientific validity.
4D Scene Persistence, Provenance & Research Object Handoff
Scene record
- Scene ID
- —
- Name
- —
- Created
- —
- Updated
- —
- Saved scenes
- 0 / 12
Provenance
- Model / demo
- —
- State digest
- —
- Selections
- —
- Release
- v0.153.0
- Object mapping
- workspace-snapshot · reference-first
Research-object handoff
- Object ID
- —
- Type
- —
- Prepared
- —
Saved scenes preserve the bounded 4D scene state, computed response surfaces and diagnostics already present in the browser, linked selections, and descriptive provenance. Loading a saved scene does not rerun computation. Research-object handoff is explicit and reference-first; it does not automatically submit to Platform Core or establish evidence, causality, significance, calibration, or scientific validity.
4D Computational Research Workspace
Workspace asset
- Asset ID
- —
- Title
- —
- Status
- —
- Revision
- —
- Parent
- —
- Updated
- —
Computational binding
- Scene digest
- —
- Compute refs
- 0
- Project assets
- 0
- Storage
- checking
- Server assets
- checking
Lineage
No lineage loaded.
Descriptive comparison
Choose two assets to compare their latest saved scene state.
Project workspace assets preserve scenes, compute-result references, immutable revisions, fork lineage, and descriptive comparisons. Loading an asset does not rerun computation. Comparisons do not establish evidence, causality, statistical significance, calibration, forecast skill, or scientific validity.
Reproducible Protocol & Computational Notebook Workspace
Reproducible protocol
Computational notebook
No draft cells yet.
Reproduction manifest
{}
Notebook compute cells are registered-method specifications executed only on explicit action through the governed Python Compute Core. Arbitrary code execution is not enabled. Protocols, notebook revisions, result references, and reproduction manifests preserve declared computational state; they do not establish evidence, causality, statistical significance, calibration, forecast skill, or scientific validity.
The front door supports bounded browser-rendered demonstrations, explicit Python Compute Core response-surface runs across user-selected X, Y and W parameters, and opt-in uncertainty, sensitivity, and seed-replication ensemble diagnostics. Illustrative browser-rendered response field. Saved project figures and governed research outputs remain distinct from this landing visualization.
Recent scientific figures
Active scientific work
LIVE / SIGNALSScientific signalsEarth, space, marine and scholarly observations
TOOLS / SPECIALISTScientific tool libraryChemistry, physics, biology, Earth systems, energy and engineering tools
PROJECT / TRACETraceability and recent activitySources → observations → evidence → models → experiments → decisions
Traceability map
Recent activity
Start a scientific record
Create a model, figure, experiment, observation, or notebook entry. Every result remains connected to the active project.
Project and Workspace Data Architecture
Project record index
| Record | Type | Collection | Status | Updated |
|---|
Schema and migration
Project checkpoints
Record relationships
Reproducible Experiment Framework
Design protocols before execution, freeze hypotheses and analysis plans, record governed runs and deviations, compare replications, and export a portable experiment bundle.
Experiment protocol
Record experiment run
Replication and reporting
Saved protocols
Run history
| Run | Protocol | Replicate | Status | Completed |
|---|
Validation, comparison, and report output
No output yet.
Parameter Studies and Design of Experiments
Generate governed experimental designs, attach them to project protocols, record observed responses, fit bounded response models, rank sensitivities, and prepare registered-method batch plans.
Study definition
Design matrix and observed response
0 runs| Design |
|---|
Saved studies
Governed output
No design generated.
Advanced experimental design & sequential experimentation
Generate bounded D-optimal or maximin designs, inspect information-matrix diagnostics, assign balanced blocks and center replicates, and propose sequential augmentation without automatically executing or stopping an experiment.
Advanced experimental design ready. The factor definition above remains the source of truth.
Optimal initial design
Initial design and observed response
Responses remain editable evidence| Run | Design |
|---|---|
| No advanced design generated. |
Optimality diagnostics
No diagnostics yet.
Sequential augmentation
Sequential proposal
| Run | Proposal |
|---|---|
| No sequential proposal yet. |
Governed record
No advanced design record yet.
Project governance
Evidence, Sources, and Research Provenance
Capture sources, quotations, assumptions, limitations, citations, and evidence links as durable project records.
Add source
Capture evidence
Assumptions and limitations
Link evidence to a project record
Sources
| Title | Authors | Year | In-text citation |
|---|
Evidence
| Source | Claim | Locator | Strength |
|---|
Research Quality and Method Review
Review scientific and computational methods against benchmark coverage, validation evidence, reproducibility, provenance, calibration, limitations, and documented reviewer decisions.
Method and review scope
Coverage and evidence
Limitations, calibration, and reviewer
Policy evaluation
Reviewer decision
Method-review registry
| Method | Risk | Benchmarks | Status | Score | Updated |
|---|
Selected review or comparison
No review selected.
Reproducible Computational Runs
Freeze governed requests, outputs, environments, package versions, checksums, warnings, and failure histories into portable project records.
Freeze a project result
Compare runs
Run manifests
Selected manifest or comparison
No run selected.
Comparison history
| Left | Right | Status | Differences | Created |
|---|
Dataset Registry and Scientific Data Management
Registered datasets
Register or import a dataset
Prepared dataset and data dictionary
Dataset metadata, profile, and lineage
External Scholarly and Data Discovery
Search governed scholarly metadata providers, deduplicate records, inspect access signals, and import selected works into the project’s evidence and provenance source library.
Scholarly search
Library profile and handoffs
Provider report
Deduplicated results
Project activity record
Scientific signal board
Climate and Earth observation map
Space telescope observations and releases
Marine biodiversity observations
Scientific dataset inspector
Import CSV or JSON
Scientific data transformation & derived variables
Reproducible scientific transformation pipeline
Transform the current working dataset through a bounded, reviewable plan. Derived variables use Lab's safe equation grammar; unit conversions use a governed compatibility catalog; every operation produces reproducible lineage hashes.
Scientific data transformation ready. Load a dataset above, then define or load a transformation plan.
Transformation plan
Supported operations: derive, filter, rename, select, drop, scale, unit-convert, cast, and explicit imputation. No arbitrary code or automatic feature engineering.
Governed dataset join
Join record
No v0.55 join yet.
Transformation evidence
Transformed data preview
Derived-variable figure
Transformation record
No v0.55 transformation yet.
Data table
Variable plot
Scientific Computing and Numerical Methods
Numerical method
Choose a governed scientific method.
- Method assumptions will appear here.
Result and provenance
Numerical output
No output.
Execution provenance
No provenance.
Raw compute response
No response.
Numerical Validation and Benchmark Library
Numerical benchmark
Catalog loading
Choose a known-answer fixture.
Tags: —
Acceptance assertions
- Assertions will appear here.
Independent browser reference
No reference loaded.
Validation result
| Status | Output | Rule | Expected | Actual | Unit |
|---|---|---|---|---|---|
| No assertion results. | |||||
Python output
No output.
Execution provenance
No provenance.
Full benchmark suite
No suite has run.
| Status | Benchmark | Method | Runtime | Error |
|---|---|---|---|---|
| No suite results. | ||||
Convergence diagnostics
Select a benchmark with a convergence series.
| Level | Parameters | Actual | Absolute error | Runtime |
|---|---|---|---|---|
| No convergence series. | ||||
Numerical Precision and Solver Governance
Governed method
Loading registry
Profiles: —
Governance result
Floating-point precision
No precision report.
Unit validation
No unit report.
Convergence and conditioning
No diagnostics.
Reference comparison
No comparison.
Numerical output
No output.
Provenance
No provenance.
Scientific Visualization for Numerical Results
Visualization specification
No specification.
Long-Running Numerical Jobs and Checkpoint Recovery
Project compute jobs
Job recovery and partial results
Chemistry laboratory
Molar mass and percent composition
Empirical formula
Molecular formula
Balance reaction
Limiting reagent and products
Theoretical yield
Reaction record
Solution concentration
Dilution
Solubility from Ksp
Strong acid or base
Weak acid or base
Buffer
Strong acid/base titration
Calorimetry
Gibbs free energy
Hess’s law
Nernst equation
Electrolysis and Faraday’s law
Arrhenius equation
Integrated rate law
Analytical calibration curve
Calibration record
Physics, electromagnetism, and particle physics laboratory
Uniform-acceleration kinematics
Projectile trajectory
Pendulum period
Spring and harmonic oscillator
Wave relation and profile
Acoustic intensity
Ideal gas state
First law and ideal limits
Fluid regime and pressure
Bernoulli pressure relation
Refraction and thin lens
Single-aperture diffraction
Photon and spectral quantities
Coulomb force and potential energy
Multiple point-charge field
Capacitor network
Magnetic fields
Lorentz force and particle orbit
Faraday induction
Electromagnetic propagation
Rectangular waveguide
Series RLC and AC power
RLC frequency sweep
First-order RC filter
Relativistic de Broglie wavelength
Particle in a one-dimensional box
Rectangular barrier tunneling
Hydrogen spectral transition
Radioactive decay and activity
Nuclear binding energy
Particle reference
Particle record
Relativistic kinematics
Invariant mass from four-vectors
Two-body decay kinematics
Detector and track reconstruction
Independent uncertainty combination
Power-law uncertainty propagation
Uncertainty-weighted mean and consistency
Physics numerical validation suite
Run deterministic benchmark cases spanning mechanics, electromagnetism, circuits, optics, nuclear physics, and particle physics. Results can be saved to the project and exported as JSON.
Physics experiment templates
Biology and computational biology laboratory
Biology numerical validation suite
Deterministic benchmark cases cover enzyme kinetics, sequence composition, translation, population genetics, diversity, qPCR, physiology, and osmotic pressure.
Biology experiment templates
Astronomy and astrophysics laboratory
Astronomy numerical validation suite
Reference cases cover coordinates, parallax, Keplerian orbits, stellar radiation, photometry, cosmology, telescope diffraction, exoplanet transits, and compact objects.
Astronomy observation and analysis templates
Materials science and characterization laboratory
Materials numerical validation suite
Reference cases cover mechanics, thermal transport, electrical resistivity, optical reflectance, diffraction, phase fractions, corrosion, composites, and microscopy resolution.
Materials characterization templates
Soil Organic Carbon Lab Foundation
Normalize soil measurements and calculate explicit fixed-depth SOC stocks by layer and profile. This foundation preserves units, depth intervals, coarse-fragment correction, fingerprints, and Carbon Project handoff provenance without inferring sequestration or credit eligibility.
Soil profile inputs
SOC workspace ready.
Fixed-depth result
Calculation record
No SOC calculation yet.
Carbon Project provenance handoff
Package the calculation as a draft model-run object compatible with the Carbon & Nature v0.4+ project-object contract. Input object IDs are linked only when you supply them.
Project packet
No project packet yet.
SOC Sampling & Field Measurement Studio
Plan governed sampling identities, normalize field samples, calculate core bulk density, retain field provenance and chain-of-custody metadata, and hand profile-ready measurements to the v0.6.0 fixed-depth SOC stock engine.
Field sample register
Sampling Studio ready.
Sampling design
Sampling plan
No plan yet.
Normalized batch
No batch yet.
Profile handoff
No handoff yet.
Carbon Project packet
No packet yet.
SOC Change & Sequestration Model
Compare matched fixed-depth SOC profiles through time, calculate explicit stock change and annualized stock-change rates, preserve temporal and spatial identity, and package the calculation as a Carbon Project model run.
Matched baseline and follow-up
SOC change workspace ready.
Matched stock-change result
Stock-change record
No comparison yet.
Carbon Project provenance handoff
Package the matched change calculation as a draft model-run object. Input object IDs are linked only when supplied; the packet does not establish verification, additionality, permanence, or credit eligibility.
Carbon Project packet
No packet yet.
SOC Spatial Variability & Uncertainty
Quantify replicate variability, Student-t confidence intervals, spatial coverage diagnostics, area-weighted stratified estimates, and paired or independent uncertainty around SOC stock change.
Replicate stock uncertainty
Stock-change uncertainty
SOC uncertainty workspace ready.
Uncertainty result
Full uncertainty record
No uncertainty analysis yet.
SOC Management Scenario Studio
Build transparent SOC management scenarios from explicit user-supplied assumptions, compare trajectories, test sensitivity to alternative rates, and preserve model provenance without converting scenarios into forecasts or carbon-credit claims.
Define a management scenario
SOC Management Scenario Studio ready.
Scenario result
Full scenario record
No scenario analysis yet.
Carbon Project packet
No packet yet.
Compare assumptions
Quick comparison uses constant annual stock-change assumptions entered as Label: rate. Results are shown in input order; Lab does not rank or recommend a scenario.
| Scenario / input | Final stock | Cumulative change | Mean annual |
|---|---|---|---|
| No comparison yet. | |||
Whole-Farm GHG Balance
Aggregate explicit greenhouse-gas emissions and removals across a farm, retain source and factor provenance, optionally integrate SOC stock change, and produce an inspectable accounting balance.
Farm GHG inputs
Format: category | emission/removal | kg CO₂e | source ref
Format: category | gas | kg gas | emission/removal | source ref. Non-CO₂ gas mass requires an explicit GWP below.
Format: category | activity value | unit | kg CO₂e/unit factor | direction | factor source | record source
Explicit GWP set
Optional SOC stock-change contribution
Optional accounting period
Whole-Farm GHG Balance ready.
Accounting result
Full GHG balance
No balance calculated yet.
Carbon Project packet
No packet yet.
Carbon MRV Method Registry
Inspect governed MRV method families, compare documentation requirements, and test whether a research packet contains the named inputs and evidence needed by a Sustainable Catalyst foundation profile.
Method registry
| Method | Type | Carbon pools | Gases | Status |
|---|---|---|---|---|
| Loading registry… | ||||
Carbon MRV Method Registry ready.
Method detail
Select a registry row to inspect its measurement basis, required evidence, uncertainty expectations, controls, and Library methodology crosswalk.
Full method profile
No method selected.
Documentation readiness
State: not checked
Readiness record
No readiness check yet.
Carbon Project monitoring-record packet
No packet yet.
Descriptive comparison
Enter 2–20 method keys. Lab preserves input order and exposes differences without ranking or recommendation.
Comparison record
No comparison yet.
MRV Protocol Builder
Turn an explicitly selected MRV foundation method into a versioned monitoring protocol scaffold with scope, responsibilities, measurement steps, QA/QC, uncertainty, data provenance, schedule, reporting, and change control.
Protocol identity & monitoring frame
Method documentation
Method-derived template
No method template loaded.
Governed protocol sections
MRV Protocol Builder ready.
Protocol readiness
Unresolved items
No validation yet.
Validation record
No validation yet.
Protocol record
Built protocol
No protocol built yet.
Carbon Project monitoring-record packet
No packet yet.
Monitoring Plan & Sampling Designer
Translate a selected MRV method and protocol into explicit monitoring campaigns, sampling-frame assumptions, precision-based sample-size planning, stratified allocation, depth design, QA/QC, and a provenance-ready monitoring record.
Plan identity & monitoring frame
Sampling design
Precision-based sample-size planning
Supply your own expected variability and precision target. No defaults are inferred.
Planning n: —
Sample-size record
No sample-size calculation yet.
Stratified allocation
Allocation record
No allocation yet.
Method documentation & evidence
Method-derived template
No method template loaded.
Monitoring Plan & Sampling Designer ready.
Plan readiness
Unresolved items
No validation yet.
Validation record
No validation yet.
Monitoring-plan record
Built plan
No monitoring plan built yet.
Carbon Project monitoring-record packet
No packet yet.
MRV Uncertainty & Detection Engine
Build explicit uncertainty budgets, quantify whether observed change exceeds a user-specified detection threshold, and plan sample sizes for minimum detectable change without hidden confidence, coverage, or power defaults.
Assessment identity
Uncertainty budget
Combined standard uncertainty: —
Budget record
No budget yet.
Observed-change detection
Threshold: — · Detected: —
Detection record
No detection calculation yet.
Detection-oriented sample-size planning
Planning n: —
Planning record
No planning calculation yet.
Evidence, review & project handoff
MRV Uncertainty & Detection Engine ready.
Assessment readiness
Unresolved items
No validation yet.
Validation record
No validation yet.
Assessment record
Built assessment
No assessment built yet.
Carbon Project model-run packet
No packet yet.
Verification Evidence Ledger
Assemble auditable MRV evidence, link each item to explicit protocol requirements, preserve internal review state, and build a deterministic hash chain without turning record integrity into a claim of external verification.
Ledger identity
Explicit evidence requirements
No evidence requirement is inferred. Carry requirements forward from the governing protocol and monitoring plan.
Evidence entries
Supported entries include observations, samples, chain-of-custody records, lab results, QC, calculations, model runs, monitoring records, protocols, methodology references, documents and images.
Provenance & project handoff
Verification Evidence Ledger ready.
Ledger readiness
Root hash: —
Unresolved items
No validation yet.
Validation record
No validation yet.
Chain check
No chain check yet.
Ledger record
Built ledger
No ledger built yet.
Carbon Project verification-record packet
No packet yet.
MRV Reporting & Audit Packets
Assemble a structured MRV report, preserve declared metric lineage, link the protocol, monitoring plan, uncertainty assessment and verification ledger, and package the result for internal audit preparation without claiming external verification or auditor approval.
Report identity
Governed component references
Internal report sections
Eight Sustainable Catalyst internal sections are expected for internal-review readiness. They are not represented as an external program or methodology template.
Reported metrics
The reporting layer preserves supplied values and lineage. It does not silently recalculate them.
Deviations & corrective actions
Open deviations block internal-report readiness. Resolved items remain visible.
Audit-packet artifact index
Build, validate & hand off
MRV Reporting & Audit Packets ready.
Readiness
Report fingerprint: —
Unresolved items
No validation yet.
Validation record
No validation yet.
Structured records
Built report
No report built yet.
Audit-preparation packet
No audit packet yet.
Carbon Project verification-record packet
No project packet yet.
Earth, climate, ocean, and marine systems laboratory
Earth systems numerical validation suite
Reference cases cover plate motion, seismic moment, atmospheric humidity, CO₂ forcing, catchment runoff, ocean waves, tsunami travel, marine diversity, remote-sensing indices, hazard recurrence, atmospheric carbon, and soil carbon.
Earth systems investigation templates
Energy and engineering laboratory
Energy and engineering numerical validation suite
Reference cases cover energy closure, solar, wind, hydro, storage, three-phase power, thermal conduction, hydrogen production, discounted cash flow, levelized cost, component reliability, and expected energy not served.
Energy and engineering study templates
Electrical, electronics, and embedded systems
Embedded device profile
Logic-level interface validation
Firmware starter artifacts
#include <Arduino.h> // Select a template to inspect starter source.
Validation suite output
Run the deterministic benchmark suite to create a hardware-validation record.
Mechanical and thermal engineering
LAB/CIVIL-INFRASTRUCTURE
Civil engineering and infrastructure systems
Auditable screening calculations for structures, soils, drainage, transportation, water systems, lifecycle cost, risk, reliability, resilience, and embodied carbon.
LAB/ARCHITECTURE-BUILDING
Architecture and building performance
Auditable calculations for building geometry, envelope performance, solar and daylight systems, ventilation and indoor environmental quality, HVAC energy, water, carbon, acoustics, and passive resilience.
LAB/URBAN-PLANNING-SPATIAL
Urban planning and spatial systems
Auditable calculations for land use, density, accessibility, mobility, spatial networks, GIS indicators, public services, equity, resilience, and urban scenarios.
LAB/SUSTAINABLE-CITIES-RESILIENCE
Sustainable cities and urban resilience
Auditable calculations for urban resource flows, climate mitigation, adaptation, critical infrastructure continuity, equity, social resilience, governance, and integrated city scenarios.
LAB/COMPARATIVE-ECONOMICS-DEVELOPMENT-SYSTEMS
Comparative economics and development systems
Auditable calculations for national accounts, economic growth, productivity, convergence, trade, structural transformation, labor markets, poverty, inequality, human development, public finance, development finance, and resilience scenarios.
LAB/AEROSPACE-ENGINEERING-FLIGHT-SYSTEMS
Aerospace engineering and flight systems
Auditable calculations for atmosphere and aerodynamics, flight mechanics, aircraft performance, stability and control, propulsion integration, structures, loads, aeroelastic screening, navigation, mission analysis, and flight-system reliability.
LAB/ROCKET-PROPULSION-SPACEFLIGHT
Rocket propulsion and spaceflight
Auditable educational and preliminary-design calculations for rocket performance, nozzle flow, launch-vehicle staging, ascent dynamics, orbital mechanics, spacecraft power and thermal systems, communications, mission budgets, and reliability.
LAB/MICROBIOLOGY
Microbiology laboratory
Auditable calculations for microbial growth, continuous culture, enumeration, microscopy, environmental microbiology, antimicrobial and disinfection screening, microbial ecology, and laboratory quality control.
LAB/CIRCULAR-ECONOMY-INDUSTRIAL-ECOLOGY
Circular economy and industrial ecology
Auditable calculations for material-flow accounting, circular products, waste prevention and recovery, industrial symbiosis, lifecycle footprints, resource productivity, and circular transition scenarios.
Universal visualization and export studio
Analysis and audit metadata
{}Dimensional scene and polytope viewer
Render true three-coordinate scenes or four-coordinate models projected 4D → 3D → 2D. Current results use direct numeric fields when available; otherwise the viewer labels the output as a normalized result-space glyph rather than physical geometry.
Scene and projection metadata
{}PDF reports and Decision Studio handoff
Report contract and audit metadata
{}Universal code switcher and multi-language execution
Execution result
Method contract
Cross-language execution
Only runtimes reported as available by the Render service are executable. Source-only languages remain viewable and downloadable.
Execution job record
No queued job is active.
Science and engineering tools
Raw data and processing
Peak detection
Spectrum and analytical results
Processing history and numerical output
Spectrometric calibration
Experiment registry
Evidence, hypotheses, and decisions
Evidence inbox
Hypotheses
Decisions
Research and laboratory notebook
Data-connected documentation
Workspace backup, restore, and reset
Back up the workspace
Download every project, record, chart, analysis packet, note, observation, and compatible preference.
The ZIP contains workspace.json, project JSON files, notebook Markdown, observations CSV, visualization records, analysis packets, a manifest, and restore instructions.
Restore a backup
Selective reset and factory reset
Last deletion receipt
No reset has been completed in this session.
Scientific source registry
System and connector status
Components
Release Integrity
LAB/BIOCHEMISTRY
Biochemistry and molecular analysis
Auditable calculations for biomolecule quantification, proteins, enzyme kinetics, nucleic acids, molecular binding, buffers, spectroscopy, separations, and laboratory quality control.
LAB/BIOTECHNOLOGY/BIOPROCESS
Biotechnology and bioprocess engineering
Auditable calculations and process simulations for cell growth, reactor balances, feed strategy, continuous culture, oxygen transfer, mixing, scale-up, production, and downstream recovery.
LAB/BIOMEDICAL/BIOSIGNALS
Biomedical engineering and biosignals
Non-clinical calculations and waveform analysis for acquisition design, ECG, PPG, respiration, EMG, EEG, filtering, and signal-quality review.
LAB/GENETICS/GENOMICS
Genetics, genomics, and sequence analysis
Deterministic sequence analysis, visualization, validation, dataset manifests, and tamper-aware provenance for research workflows.
LAB/DATA/INSTRUMENTATION
Laboratory data and instrumentation
Instrument, sensor, sample, run, calibration, maintenance, measurement, and custody records with normalized ingestion and auditable manifests.
VISUALIZE / GRAPH STUDIO / COMPETING PATHS v0.135.11.0
Scientific Visualization Workspace
Graph Studio now compares multiple bounded provenance paths between two governed research objects. Researchers can inspect alternative declared relationship chains, switch the active path without rebuilding the graph, review explicitly stored supporting or contradicting edge context, and hand the comparison to Project Workspace without asking the Lab to choose a preferred explanation.
Graph Studio initializing scientific context…
Interactive scientific canvas
Presentation changes remain separate from the underlying scientific record.
Project figure library
0 FIGURESScientific graph specification
No figure rendered.
PROJECT / MODEL STUDIO / v0.46.0
Response Surfaces, Optimization & Design-Space Exploration
Fit auditable quadratic response surfaces from bounded experimental factors, inspect coefficient and lack-of-fit evidence, explore constrained design spaces, identify governed optima, and carry the results into the shared interactive scientific graph, dynamic-system, uncertainty, and publication workflows.
Model Studio loading…
Model definition
Safe grammar: + − × ÷ ^ and registered functions including exp, log, sqrt, sin, cos, tan, min, max and abs.
Algebraic equation previews remain governed by the v0.42 safe grammar. Dynamic-system derivative expressions retain the same v0.45 whitelist. Registered calibration diagnostics remain separate; arbitrary code and arbitrary algebraic-form fitting remain disabled.
Data binding & scientific graph
Publication figure
These settings are embedded into exported figures. The graph toolbar provides reset, zoom, pan, SVG, 2× PNG, CSV, JSON, and accessible-table controls.
Evaluated rows
No evaluation yet.
Scientific graph specification
No graph specification yet.
Governed handoffs
Lab ↔ Workbench model exchange
Move the same governed computational model between scientific validation in Lab and interactive exploration in Workbench. The handoff preserves scientific definition and provenance without enabling arbitrary code.
Shared Lab ↔ Workbench model contract loading…
Shared computational model handoff
No Workbench handoff prepared.
Reproducible model package
Freeze the shared computational model together with its dataset reference or snapshot, methods, results, figures, environment, assumptions, limitations, and provenance. Packages are content-hashed, independently verifiable, portable as JSON or a research ZIP, and can be registered as immutable model versions.
Registry versions are immutable. If scientific content changes after registration, choose a new semantic version.
Reproducible model package workspace loading…
Package record and verification
No reproducible model package built.
Model contract
No model contract yet.
Advanced statistical modeling & generalized regression
Fit multivariable Gaussian regressions, robust and regularized models, binomial and Poisson generalized linear models, and cubic splines. Validation, coefficient evidence, scientific figures, and candidate comparison stay connected to the active Lab project and reproducible-model workflow.
Advanced statistical modeling loading…
Scientific boundary. Model coefficients describe associations under the declared model and data. Lab does not infer causality, automatically select features, execute arbitrary formulas, or attach classical p-values to regularized or Huber estimates.
Statistical model / validation record
No advanced statistical result yet.
Bayesian inference, posterior diagnostics & posterior predictive modeling
Estimate posterior distributions for Gaussian, binomial-logit, and Poisson-log regression models with explicit priors, multi-chain sampling, diagnostic review, credible intervals, and posterior-predictive checks. Bayesian evidence remains connected to the active project, Graph Studio, and reproducible-model packages.
Priors
Sampler
Bayesian modeling loading…
Scientific boundary. Posterior claims are conditional on the declared model, data, priors, and sampler. Split-R̂, ESS, MCSE, acceptance rates, trace plots, and posterior-predictive checks are review evidence—not an automatic convergence certificate. Lab does not select priors automatically, infer causality, execute arbitrary code, or claim that a posterior distribution proves a real-world mechanism.
Posterior summary
Chain diagnostics
Bayesian model / posterior record
No Bayesian result yet.
Diagnostics & model evidence
Fit registered model forms against observed response values, inspect residual behavior, estimate out-of-sample performance, and compare alternatives using the same deterministic validation policy.
Diagnostics / validation record
No diagnostics record yet.
Scientific model comparison
Select two or more registered forms. Lab evaluates each candidate with the same folds, repeats, seed, response, features, and dataset, then ranks primarily by mean validation RMSE with AICc and BIC as tie-breakers.
Model comparison record
No model comparison yet.
Response surfaces, optimization & design-space exploration
Fit a full second-order response surface in coded factor space, including linear terms, quadratic curvature, and all two-factor interactions. Lab reports coefficient uncertainty, design-matrix conditioning, information criteria, and a replicated-point lack-of-fit test when the experimental design provides pure-error degrees of freedom.
Explore bounded design space
Bounded optimization
Response-surface / design-space record
No response-surface record yet.
Dynamic systems & ODE parameter estimation
Build deterministic coupled ordinary differential equation systems with declared states, bounded parameters, explicit initial conditions, and a governed solver. Derivative expressions reuse the safe v0.42 scientific grammar; no Python or arbitrary code is executed.
Estimate parameters from observations
Bind observations by the independent-variable symbol and one or more state symbols. Lab estimates only parameters not marked fixed, respects configured bounds, and reports local Jacobian identifiability evidence.
Dynamic-system / estimation record
No dynamic-system record yet.
Dynamic Systems II — events, regimes, bifurcation & phase analysis v0.54.0
Extend the governed ODE definition above with safe event functions, evidence-backed scheduled regime changes, bounded numerical parameter sweeps, and two-state autonomous phase-plane diagnostics. These tools remain deterministic and declarative; no arbitrary Python or executable callbacks are accepted.
Dynamic Systems II ready. Configure the base ODE above, then add advanced analysis only where scientifically appropriate.
Numerical bifurcation scan
Advanced two-state phase analysis
Dynamic Systems II record
No v0.54 analysis yet.
System Dynamics, Feedback Loops & Stock-Flow Modeling v0.86.0
Build Meadows-style causal-loop and stock-flow models with explicit polarity, delays, safe equations, Euler/RK4 simulation, and structural leverage indicators. Causal links and leverage meaning remain human-declared; Lab detects structure but does not invent causal relationships or rank policy leverage automatically.
System dynamics workspace ready.
System dynamics record
No v0.86 system analysis yet.
PROJECT / MODEL CALIBRATION / v0.30.2
Scientific model calibration and validation
Fit registered scientific model forms to project datasets, reserve holdout observations, inspect residuals and parameter uncertainty, compare models, and preserve calibration provenance.
Model calibration loading…
Calibration output
Run a governed calibration to inspect parameters, confidence intervals, residuals, holdout metrics, and provenance.
PROJECT / WORKFLOW ORCHESTRATION / v0.32.1
Workflow Checkpoints, Conditional Execution, and Partial Recovery
Build dependency-aware scientific workflows with safe declarative conditions, durable checkpoint history, resumable node context, and lineage-preserving recovery runs that reuse successful work while restarting only affected branches.
Integrated Scientific Research Beta v0.60.0
One coherent research journey across the Lab
Assess whether the active project has the metadata evidence needed to move coherently from data and models through validation, uncertainty, experiments, figures, workflows, reproducibility, compute hardening, and scientific audit. v0.60 evaluates evidence only; it does not certify science or authorize publication.
Integrated beta loading…
Integrated research journey
No beta journey assessment yet.
Readiness / packet evidence
No integrated beta evidence has been generated yet.
The beta packet contains metadata, counts, contract hashes, and evidence hashes only. Raw datasets, credentials, and scientific inputs are excluded. Human review remains mandatory.
End-to-End Scientific Study & Research Project Lifecycle v0.61.0
Turn the integrated Lab project into one governed scientific study
Define the research question and protocol, connect existing project evidence, review each required lifecycle stage, preserve conclusions and limitations, and build a tamper-evident study packet. v0.61 records scientific project state; it does not certify science, infer causality, authorize physical experiments, or publish automatically.
Scientific study lifecycle loading…
Scientific study definition
Study lifecycle and human review
No lifecycle assessment yet.
Saved project studies
Lifecycle / packet evidence
No scientific study lifecycle evidence generated yet.
Study packets contain study metadata, stage states, evidence references/hashes, and review hashes. Raw datasets, credentials, arbitrary code, and automatic scientific claims are excluded.
Scientific Claims, Evidence Matrix & Conclusion Traceability v0.62.0
Trace every scientific conclusion back through explicit evidence
Record claims as researcher-authored statements, link supporting and contradicting evidence, preserve uncertainty and limitations, and connect final conclusions to reviewed claims. v0.62 evaluates traceability and evidence completeness; it does not infer claims, certify scientific validity, or authorize causal conclusions.
Scientific claims traceability loading…
Scientific claim
Conclusion trace
Project evidence catalog
Click an evidence record to copy its reference. Only metadata, IDs, titles, and existing hashes are cataloged here; raw scientific values are not sent to the traceability evaluator.
Saved claims
Saved conclusions
Claim × evidence matrix
| Claim | Type | Gate | Primary support | Contradictions | Uncertainty / limits | Conclusions |
|---|---|---|---|---|---|---|
| No claims evaluated. | ||||||
Conclusion traceability
| Conclusion | Gate | Claims | Review |
|---|---|---|---|
| No conclusions evaluated. | |||
Matrix / packet evidence
No scientific claim traceability evidence generated yet.
Contradicting evidence is retained rather than suppressed. Figure links provide context/presentation lineage but do not substitute for primary support. Causal claims are blocked when the active v0.61 study design is not experimental or mixed.
Scientific Literature, Citation Graph & Source-to-Claim Provenance v0.63.0
Extend claim traceability outward to published and archival sources
Register bibliographic metadata, review source applicability, connect literature to v0.62 scientific claims, preserve contradictory findings, and record source-to-source citation relations. v0.63 traces provenance; it does not rank authority, determine truth, verify retractions automatically, or fetch literature during evaluation.
Scientific literature provenance loading…
Literature source record
Saved literature sources
Source → claim provenance
Citation graph edge
Claim × literature provenance
| Claim | Gate | Sources | Reviewed | Contradictions | Unresolved |
|---|---|---|---|---|---|
| No literature provenance evaluated. | |||||
Source-to-claim links
| Claim | Role | Source | Locator |
|---|---|---|---|
| No links saved. | |||
Citation graph
| From | Relation | To |
|---|---|---|
| No citation edges saved. | ||
Literature / citation provenance evidence
No literature provenance evidence generated yet.
Full-text documents are not copied into this packet. Citation counts are not authority scores. Retraction/correction status is declared metadata unless a separate governed connector verifies it. Contradicting and non-replication sources remain visible.
Replication, Systematic Evidence Synthesis & Meta-Analysis v0.64.0
Synthesize reviewed literature without collapsing disagreement
Register aggregate study effect estimates from v0.63 literature sources, evaluate replications, calculate fixed- or random-effects summaries, preserve heterogeneity, and attach the result back to the scientific claim set. v0.64 summarizes evidence; it does not determine scientific truth.
Evidence synthesis loading…
Synthesis protocol
Human synthesis review
Aggregate study effect estimate
Saved protocols
Study effects
| Effect | Source | Metric | Estimate | SE | Replication of |
|---|---|---|---|---|---|
| No aggregate effect estimates saved. | |||||
Replication assessment
| Replication | Original | Gate | Δ z |
|---|---|---|---|
| No replication links recorded. | |||
Synthesis / meta-analysis evidence
No evidence synthesis generated yet.
A pooled estimate is a model-based summary, not a truth score. I² and τ² remain explicit; leave-one-out sensitivity and replication disagreement remain visible. Study quality is declared/reviewed upstream rather than inferred from citation counts or journal prestige.
Scientific Evidence Grading, Contradiction Analysis & Consensus Boundaries v0.65.0
State what the evidence supports — and where agreement stops
Grade the evidence state for reviewed v0.62 claims using v0.63 literature provenance and v0.64 synthesis/replication evidence. Contradictions, replication disagreement, heterogeneity, source caution, and scope limitations remain explicit. v0.65 defines a reviewable evidence boundary; it does not calculate a truth score or certify scientific consensus.
Evidence grading loading…
Evidence boundary definition
Human boundary review
Evidence grading rules
Substantial evidence requires a reviewed synthesis, multiple reviewed supports, at least one directionally consistent replication, no material contradiction, and no high heterogeneity. Moderate and limited states remain useful evidence but cannot become an unqualified bounded consensus automatically.
Saved assessments
Current gate
needs-review
Claim × evidence grading matrix
| Claim | Grade | Support | Contradictions | Replication C/D | Boundary candidate |
|---|---|---|---|---|---|
| No evidence grading assessment yet. | |||||
Contradiction analysis
| Claim | State | Factors | Qualified interpretation? |
|---|---|---|---|
| No contradiction analysis yet. | |||
Evidence grade / consensus boundary evidence
No v0.65 evidence grading evidence generated yet.
A consensus boundary is a scoped human-reviewed statement about the current evidence record, not a declaration of truth. Contradicting sources, non-replications, heterogeneity, limitations, source cautions, unresolved references, and qualified reviews remain in the packet.
Competing Hypotheses & Scientific Argumentation v0.66.0
Compare rival explanations without turning evidence into an automatic winner
Define competing hypotheses, state predictions before interpretation, map governed evidence to each alternative, design discriminating tests, preserve falsifying challenges, and record explicit human judgments. v0.66 structures scientific argumentation; it does not prove, automatically falsify, or rank hypotheses by a hidden truth score.
Scientific argumentation loading…
Argument case
Human argument review
Competing hypothesis
Saved hypotheses
Hypothesis evidence mapping
| Hypothesis | Evidence | Relation | Test |
|---|---|---|---|
| No hypothesis evidence links. | |||
Discriminating test
| Test | Alternatives | Expected outcomes | Status |
|---|---|---|---|
| No discriminating tests. | |||
Human hypothesis review
Current gate
needs-review
Argumentation evidence
No v0.66 scientific argumentation evidence generated yet.
Hypothesis × evidence assessment
| Hypothesis | State | Predictions | Support | Contradictions | Falsifying challenges | Human review |
|---|---|---|---|---|---|---|
| No hypothesis assessment yet. | ||||||
Discrimination analysis
| Test | Hypotheses | Distinct predictions | Discriminates? |
|---|---|---|---|
| No discriminating tests evaluated. | |||
Scientific argumentation is an explicit map of alternatives and evidence. A retained or provisionally preferred hypothesis remains open to revision; rejected alternatives, contradictory evidence, falsifying candidates, failed discriminating tests, and unresolved uncertainty remain in the provenance packet.
Causal Inference & Quasi-Experimental Methods v0.67.0
Estimate causal contrasts without hiding identification assumptions
Define a governed quasi-experimental design, state the estimand and assignment mechanism, record identification assumptions, add aggregate effect estimates, inspect method-specific diagnostics and sensitivity evidence, and preserve explicit human review. Supported methods include matching, weighting, difference-in-differences, interrupted time series, and regression discontinuity.
Causal inference loading…
Causal design
Human causal review
Aggregate causal estimate
Project causal records
No v0.67 causal records saved.
Method diagnostic / sensitivity check
Current gate
needs-review
Causal inference evidence
No v0.67 causal inference evidence generated yet.
Identification assumption coverage
| Required assumption | State |
|---|---|
| No assumption assessment yet. | |
Method diagnostic coverage
| Required diagnostic | State |
|---|---|
| No diagnostic assessment yet. | |
v0.67 accepts aggregate estimates and governed diagnostic metadata only. Raw participant-level data, treatment/outcome vectors, hidden assumption satisfaction, automatic causal certification, arbitrary code, and network fetching during evaluation are excluded.
Hierarchical, Multilevel & Cross-Study Modeling v0.68.0
Model variation across groups and studies without erasing structure
Fit governed aggregate hierarchical models with explicit partial pooling, random-intercept cluster pooling, random-slope/meta-regression structure, between-unit heterogeneity, shrinkage diagnostics, and human-reviewed generalization boundaries. v0.68 works from aggregate unit/study estimates and reported uncertainty; participant-level records are outside this release boundary.
Hierarchical modeling loading…
Hierarchical model definition
Human multilevel-model review
Aggregate unit / study estimate
Project multilevel records
No v0.68 multilevel records saved.
Aggregate unit estimates
| Unit | Cluster | Source | Estimate | SE | Moderator |
|---|---|---|---|---|---|
| No v0.68 aggregate unit estimates. | |||||
Current gate
needs-review
Hierarchical modeling evidence
No v0.68 hierarchical modeling evidence generated yet.
Shrinkage / model-fit diagnostics
| Unit | Observed | Shrunken / predicted | Pooling weight / residual |
|---|---|---|---|
| No shrinkage/residual diagnostics yet. | |||
Interpretation boundary
Report τ² and I² alongside pooled estimates. Treat large shrinkage as a model consequence, not proof that unit differences are noise. Random-slope/meta-regression estimates apply over the observed moderator range unless a researcher explicitly justifies otherwise.
v0.68 accepts aggregate unit/study estimates and their uncertainty only. Raw participant records, automatic generalization, ecological inference, hidden model selection, automatic causal proof, arbitrary code, and network fetching during evaluation are excluded.
Scientific Theory & Conceptual Model Workspace v0.69.0
Connect constructs, mechanisms, hypotheses, equations, claims, and evidence into an explicit theory
Build governed conceptual and mechanistic models with defined constructs, typed relations, testable predictions, falsification conditions, scope boundaries, and explicit human review. v0.69 represents a scientific theory as a challengeable model of explanation; conceptual coherence is not scientific proof.
Scientific Theory workspace loading…
Theory definition
Human theory review
Construct
Project theory records
No v0.69 theory records saved.
Relation / mechanism
Testable prediction
Conceptual graph
| Source | Relation | Target | Mechanism hash |
|---|---|---|---|
| Build the conceptual graph to inspect theory relations. | |||
Current gate
needs-review
Theory evaluation / packet evidence
No v0.69 scientific theory evidence generated yet.
v0.69 stores conceptual metadata, hashes, definitions, relations, predictions, evidence references, falsification conditions, scope, limitations, and human review. Raw datasets, automatic theory proof, automatic causal certification, universal generalization, arbitrary code, and network fetching during evaluation are excluded.
Research Questions, Hypothesis Registry & Preregistration v0.70.0
Freeze the scientific plan before results are interpreted
Register explicit research questions, hypotheses, outcomes, analysis rules, exclusions, stopping rules, information-size rationale, and sensitivity plans. Freeze the registered metadata before outcome results are inspected, then disclose post-freeze deviations without rewriting the original plan.
Preregistration workspace loading…
Preregistration plan
Human preregistration review
Research question
Project registration records
No v0.70 registration records saved.
Hypothesis registry
Post-freeze deviation log
Do not edit the frozen snapshot to make later choices look preregistered. Record the change here instead.
Current gate
needs-freeze
A freeze records hashes of the registered questions, hypotheses, outcomes, analysis plan, exclusions, stopping rule, and other prespecified boundaries. Changes after freeze must be disclosed as deviations.
Preregistration evaluation / packet evidence
No v0.70 preregistration evidence generated yet.
v0.70 stores governed research metadata, hashes, freeze snapshots, deviation records, scope, limitations, and human review. Raw datasets, participant-level records, automatic hypothesis validation, automatic post-hoc preregistration, automatic outcome reclassification, arbitrary code, and network fetching during evaluation are excluded.
Beta Field Diagnostics, Integration Soak & Runtime Repair v0.60.1
Exercise the integrated beta before adding more science
Capture a metadata-only runtime snapshot, localize cross-studio failures, and run a bounded four-cycle WordPress ↔ compute integration soak. The diagnostic layer reports repair evidence and failure domains; it does not mutate scientific state or send external telemetry.
Beta field diagnostics loading…
Endpoint soak
No soak result yet.
Failure domains
Repair evidence
- Run the integration probe to localize a failure.
Diagnostic evidence
No beta field diagnostic evidence generated yet.
The soak is bounded and user initiated. Runtime metadata, endpoint status/latency, hashes, and failure categories may be retained; raw datasets, credentials, model inputs, automatic repairs, and external telemetry are excluded.
Scientific Workflow Composer v0.57.0
Compose reproducible scientific pipelines
Connect governed Lab stages into a saved, rerunnable research workflow. v0.57 executes registered scientific adapters only; arbitrary code, external callbacks, automatic experiment execution, registry promotion, and publication remain disabled.
Workflow Composer loading…
No workflow loaded.
Stage timeline
| Stage | Kind | Status | ms | Output hash |
|---|---|---|---|---|
| No workflow run yet. | ||||
Saved project workflows
| Workflow | Stages | Hash | Action |
|---|---|---|---|
| No saved v0.57 workflows in this project. | |||
Workflow definition and run inputs
Latest run record
No workflow run yet.
Large workload & compute hardening v0.58.0
Keep scientific workflows responsive at scale
Assess workload size before execution, queue large runs without blocking the interface, reuse deterministic cached results, inspect bounded dataset windows, and request cooperative cancellation. Running scientific code is never force-terminated and arbitrary executable jobs remain disabled.
Compute hardening loading…
Recent hardened compute jobs
| Job | Operation | Status | Cache | Action |
|---|---|---|---|---|
| No hardened compute jobs yet. | ||||
Assessment / result evidence
Assess the current workflow or queue it for bounded asynchronous execution.
Security, privacy, reproducibility & scientific audit v0.59.0
Audit research evidence before preservation or external sharing
Scan the active scientific workflow for prohibited executable fields, secret/credential leakage, direct identifiers, data-minimization concerns, and missing reproducibility evidence. Findings record paths and categories only; sensitive values are not echoed into the audit report.
Scientific audit loading…
No v0.59 scientific audit has been run yet.
Audit results are decision support, not certification. Automatic publication, deletion, high-stakes decisions, and scientific certification remain disabled.
Workflow recovery loading…
Workflow definition
Conditions are declarative data only. Arbitrary Python, shell code, JavaScript, and callback URLs are rejected.
Run controls
Partial recovery
Checkpoint controls
Workflow runs
| Run | Workflow | Project | Status | Recovery generation | Updated | Action |
|---|
Definition, run state, checkpoints, and timeline
No response yet.
PROJECT / WORKFLOW AUTOMATION / v0.32.2
Scheduled and Event-Driven Research Runs
Run saved scientific workflows from durable UTC schedules or authenticated external events, with missed-run recovery, duplicate-event protection, concurrency policies, firing history, and direct workflow-run provenance.
Workflow automation loading…
Schedule definition
Manual and scheduler controls
Event ingestion
Backend event routes require normal compute authentication. Deployments may additionally require an HMAC event signature.
Saved schedules
| Schedule | Workflow | Trigger | Status | Next fire | Actions |
|---|
Automation state, firings, and event receipts
No response yet.
PROJECT / ADAPTIVE EXPERIMENTS / v0.33.1
Bayesian Optimization, Active Learning, and Resource-Aware Search
Fit an inspectable Gaussian-process surrogate to completed workflow trials, quantify predictive uncertainty, select the next experiment with governed acquisition policies, and trade expected scientific value against explicit resource costs.
Bayesian campaign engine loading…
Campaign definition
Campaign and model controls
Manual observation
Import an external, simulation, or instrument observation with both objective and resource cost so it can train the same surrogate.
Safety: the surrogate proposes only typed parameter values. It cannot execute arbitrary code, shell commands, callback URLs, or unregistered methods.
Saved campaigns
| Campaign | Workflow | Strategy | Status | Completed / budget | Best objective | Cost | Surrogate | Actions |
|---|
Campaign, surrogate, prediction, acquisition, trial, and timeline records
No response yet.
PROJECT / MODEL GOVERNANCE / v0.34.0
Scientific Model Registry and Environment Reproduction
Register immutable scientific model versions, lock runtime and dependency environments, promote reviewed versions, preserve deprecation history, and export portable reproduction manifests with cryptographic verification.
Scientific model registry loading…
Register model version
Version governance
Integrity rule: registered versions are immutable. Any change to model content, dependencies, artifacts, or execution parameters requires a new semantic version.
Capture environment
Registered model versions
| Model | Version | Type | Channel | Action |
|---|
Reproduction manifest
Registry, environment, promotion, and verification records
No response yet.
MODEL / UNCERTAINTY & SENSITIVITY / v0.53.0
Integrated Uncertainty, Sensitivity & Probabilistic Visualization
Propagate explicit input and parameter uncertainty through safe declarative Model Studio models, represent independent or evidence-backed dependent inputs, quantify output distributions and threshold probabilities, inspect dependence-aware sensitivity, and carry probabilistic figures directly into Graph Studio.
Probabilistic analysis loading…
Probabilistic visualization
Sensitivity evidence
Threshold probabilities
Dependency diagnostics
Run a study to compare the declared dependency model with sampled dependence.
Probabilistic analysis record
No probabilistic analysis yet.
PROJECT / MODEL ANALYSIS / v0.34.1
Ensemble Simulation, Global Sensitivity, and Uncertainty
Run immutable registered-model ensembles through governed workers, propagate uncertain inputs, compare weighted model members, and quantify output distributions and global parameter influence.
Ensemble and uncertainty engine loading…
Ensemble study definition
Governance: members must be immutable registered-method model versions. Lab never executes arbitrary Python, shell commands, or callbacks from a study definition.
Study controls
Operator-recorded result
Use only for validated external results or test fixtures. Normal studies reconcile results automatically from the dispatcher.
Saved studies
| Study | Status | Project | Definition | Action |
|---|
Study, evaluation, uncertainty, sensitivity, and provenance records
No response yet.
PROJECT / MODEL ACCELERATION / v0.34.2
Surrogate Models and Reduced-Order Analysis
Train reproducible surrogate models, compress high-dimensional simulation states with proper orthogonal decomposition, validate approximation error, reconstruct reduced states, and publish immutable surrogate versions into the Scientific Model Registry.
Surrogate and reduced-order engine loading…
Training definition
Modes: use surrogate for scalar response approximation, reduced-order for POD/SVD state compression, or hybrid-rom to learn parameter-to-reduced-state mappings. Definitions cannot execute arbitrary Python, shell commands, or callbacks.
Study controls
Prediction or reconstruction
Publish to model registry
Saved surrogate and ROM studies
| Study | Mode | Status | Model hash | Action |
|---|
Training, validation, prediction, reduced basis, registry, and provenance records
No response yet.
PROJECT / COLLABORATION / v0.35.0
Shared Research Projects and Team Workspaces
Create private research workspaces with explicit membership roles, single-use invitations, governed links to Lab resources, ownership continuity, durable access decisions, and a complete collaboration timeline.
Shared workspace service loading…
Workspace definition
Integration: v0.35.0 governs membership and shared resource access; v0.35.1 adds review, approvals, and sign-off in the adjacent workspace.
Workspace controls
Invite member
Accept invitation
Invitation tokens are returned only once. Send the token through an appropriate private channel.
Member governance
Transfer ownership
Link governed Lab resource
Access decision
Accessible team workspaces
| Workspace | Your role | Status | Primary project | Action |
|---|
Workspace, membership, invitation, resource, access, and activity records
No response yet.
PROJECT / COLLABORATION GOVERNANCE / v0.35.1
Review, Comments, Approvals, and Scientific Sign-Off
Run append-only scientific discussions, assign reviewers, enforce explicit approval gates, prevent stale concurrent updates, and create immutable sign-off records for governed workspace resources.
Review and sign-off service loading…
Workspace context
Create review thread
Append-only comment
Resolve or reopen thread
Review assignment
Approval request
Immutable decision
Scientific sign-off
Review, approval, decision, sign-off, and event records
Revision-bearing updates reject stale clients with HTTP 409. Decisions and sign-offs are append-only and cannot be edited or deleted.
No response yet.
PROJECT / COLLABORATION GOVERNANCE / v0.35.2
Version History, Branching, Merge, and Conflict Resolution
Preserve immutable research states, develop changes on named branches, compare snapshots, resolve path-level conflicts, and merge approved work without rewriting scientific history.
Workspace version-control service loading…
Workspace context
Create branch
Create immutable snapshot
Compare or restore
Open merge request
Resolve conflict
Reviewers may resolve path-level conflicts. Every resolution retains the base, source, target, actor, timestamp, and conflict hash.
Protected-branch approval
Protected targets require a signed v0.35.1 scientific approval before finalization.
Finalize or cancel merge
Branches, snapshots, comparisons, conflicts, approvals, and history
Snapshots and merge records are immutable. Rollback creates a new snapshot. Branch heads and merge revisions reject stale clients with HTTP 409.
No response yet.
PROJECT / ARTIFACT GOVERNANCE / v0.36.0
Scientific Artifact Repository and Data Federation
Organize immutable scientific artifacts into governed collections, verify content identities, exchange hashed manifests with trusted institutional nodes, preserve tombstones, and resolve federation conflicts without enabling unrestricted remote execution.
Workspace and collection
Register and verify artifact
Use transportArtifactId to bind a repository record to bytes already verified by Lab’s content-addressed artifact transport.
Federated source
Manifest synchronization
Federation boundary: synchronization accepts explicit manifests through authenticated Lab routes. The coordinator does not make arbitrary callbacks to submitted URLs.
Conflict resolution
Collections, artifacts, sources, sync runs, conflicts, and integrity records
No response yet.
PROJECT / INSTITUTIONAL FEDERATION / v0.36.1
Institutional Node Federation and Local-Data Execution
Register governed institutional compute nodes, describe data that must remain local, issue signed registered-method envelopes, and retain attested aggregate results without centralizing confidential or restricted datasets.
Institutional node
The node secret is returned once during registration. Store it only in the institutional node runtime, never in WordPress or a browser.
Local data asset
Only metadata, schema hashes, classifications, and policy records are centralized. Restricted data bytes stay on the institutional node.
Governed local execution
Execution boundary: requests reference registered Lab methods and local data IDs. Arbitrary Python, shell commands, executable expressions, and unrestricted callbacks are rejected.
Cancel local execution
Completed results require a node-authenticated attestation containing the result hash, data-access digest, environment hash, and policy-approved summary.
Nodes, local data assets, execution envelopes, attestations, and federation history
No response yet.
PROJECT / FIELD RESEARCH / v0.36.2
Offline Field Research and Edge Synchronization
Create sealed field-work packages, enroll governed edge devices, synchronize observations after disconnected work, and resolve concurrent edits without moving restricted institutional data into the browser or field package.
Edge device
The device secret is shown once and belongs only in the encrypted edge runtime. It is never stored in WordPress or exposed to browser-side synchronization.
Offline work package
Packages contain protocols, forms, method IDs, hashes, and references—not restricted data bytes, arbitrary code, shell commands, or callback URLs.
Synchronization oversight
Edge runtimes use signed batches, cursor-based deltas, one-time resume tokens, duplicate change IDs, and immutable device provenance. Browser controls provide oversight only.
Conflict resolution
Available resolutions preserve local, accept edge, retain both, or dismiss the conflict. Reconciliation creates a new event rather than silently rewriting field history.
Devices, packages, synchronization sessions, conflicts, and field provenance
No response yet.
PROJECT / REPRODUCIBILITY & PUBLICATION / v0.37.0
Reproducibility Packages & Research Publication Studio
Assemble workspace-governed research bundles, verify immutable manifests, render publication-ready outputs, and publish only after scientific sign-off.
Connecting to the publication studio…
Reproducibility package
Sealed packages are immutable. Raw restricted data, executable code, shell commands, secrets, and unrestricted callbacks are rejected.
Research publication
Scientific publication gate
Publication requires a verified sealed package, a completed approval, and an immutable scientific sign-off bound to the same workspace resource.
Packages, verification receipts, publication outputs, and provenance
No response yet.
PROJECT / RESEARCH ASSEMBLY / v0.37.1
Manuscript, Report, Notebook & Methods Assembly
Build structured research documents from reusable sections and sealed reproducibility packages, generate methods narratives, render safe output-only notebooks, and preserve immutable revision lineage.
Connecting to the assembly studio…
Workspace and section library
Research assembly
Methods and outputs
Notebook exports contain narrative and stored result cells only. They never contain executable code.
Revision lineage
Assembly records, validation, renders, and lineage
No response yet.
PROJECT / PUBLIC VERIFICATION / v0.37.2
Public Reproduction & Verification Portal
Publish safe immutable reproduction records, expose public manifests, issue nonce challenges, and verify independent hash evidence without executing submitted code or centralizing restricted data.
Connecting to the public verification portal…
Workspace record
Public manifest and challenge
The public view excludes workspace member identities, secrets, executable content, and restricted dataset bytes.
Independent verification evidence
Receipt lookup
Withdrawal
Withdrawal preserves a public tombstone and complete verification history; it does not hard-delete the record.
Records, manifests, challenges, receipts, and provenance
No response yet.
INSTITUTION / CONTROL PLANE / v0.39.0
Institutional Administration, Identity, and Governance
Administer institutions, organizational units, credential-free human and service principals, role bindings, workspace classifications, retention policies, approval gates, and hash-chained governance decisions.
Connecting to the institutional governance service…
Institution and governance context
This release stores identity and authority records, not passwords, API secrets, or SSO tokens. Credential hardening is reserved for v0.39.1.
Create an institution
Organizational units
Human and service principals
Service principals are credential-free identity records in v0.39.0.
Institutional role bindings
Retention policy
Workspace governance profile
Policy evaluation
Approval request and decision
Institutional governance response
No response yet.
PLATFORM / CONTROLLED BETA / v0.40.0
Connected Scientific Research Platform Beta
Operate controlled institutional cohorts, guide complete research journeys, collect privacy-minimized opt-in telemetry, triage feedback and limitations, connect support pathways, and evaluate beta release readiness across the Lab’s governance, security, interoperability, recovery, and validation layers.
Connecting to connected platform beta operations…
Institutional beta cohort
Institution onboarding journey
Guided end-to-end research project
Privacy-minimized beta telemetry
Actor and workspace identifiers are hashed. Raw identifiers, secrets, restricted data, and research payloads are rejected.
Feedback and known limitations
Support and incident pathway
Beta release-readiness gate
Expansion is blocked when a required platform component is degraded or a critical feedback, limitation, or support record remains unresolved.
Beta operations response
No response yet.
INTERFACE / ACCESSIBILITY / MOBILE / OFFLINE / v0.40.1
Accessible Mobile and Offline Research Workspace
Audit the active Lab interface at phone, tablet, and desktop widths, apply accessibility preferences, retain browser-local project snapshots, queue disconnected work, reconcile conflicts explicitly, and opt into a restricted-data-safe offline shell.
Checking connection and interface state…
Responsive accessibility audit
Runs a browser audit for overflow, 44 px targets, keyboard behavior, accessible names, live regions, dialogs, tables, visualizations, reduced motion, forced colors, and text zoom.
Accessibility and data-use preferences
Opt-in offline shell
Caches static Lab assets. A public Lab page is cached only after this explicit action; authenticated pages are not cached. Offline research payloads stay in IndexedDB on this device.
Browser-local project snapshot
Governed offline snapshot metadata
Disconnected operation queue
Explicit synchronization reconciliation
Interface and offline operations response
No response yet.
STABLE / CONNECTED / RESEARCH / COMPUTE / v1.0.0
Connected Scientific Research and Compute Platform
Register stable public contracts, certify upgrades, declare support lifecycles, attest production and incident readiness, and issue evidence-bound general-availability certifications for the complete connected research platform.
Connecting to the stable platform…
Stable contract registry
Support lifecycle
Upgrade certification
Production readiness
Incident readiness
General availability
Stable platform response
No response yet.
RELEASE / MIGRATION / COMPATIBILITY / v0.40.2
Migration, Compatibility, and Public Release Hardening
Assess supported upgrades, publish compatibility matrices, retain deprecation guidance, prove clean installs and rollback procedures, and evaluate release-candidate evidence without force-pushing or overwriting production files through the API.
Connecting to public release hardening…
Compatibility matrix
Migration dry run
Deprecation registry
Clean-install report
Rollback proof
Release-candidate gate
Release-hardening response
No response yet.
SYSTEM / RELEASE VALIDATION / v0.39.3
Performance, Load, and Chaos Validation
Measure latency percentiles and throughput, enforce explicit performance budgets, inject bounded failures into isolated resources, and produce evidence-bounded capacity reports without destructive production chaos or external traffic.
Connecting to performance validation…
Load validation
Validation workloads use temporary or dedicated validation resources. They do not mutate active research records.
Safe chaos scenario
Available scenarios: database lock, storage latency, worker termination, network timeout simulation, and partial-write rejection. Production chaos remains disabled.
Capacity evidence
Reports summarize validated evidence only and never claim production sizing without production-equivalent tests.
Validation response
No response yet.
SYSTEM / RECOVERY OPERATIONS / v0.39.2
Multi-Instance Operations, Backup, Migration, and Disaster Recovery
Identify this Lab instance, create consistent signed backups, verify archives, stage non-destructive restores, execute idempotent migration plans, transfer governed recovery bundles between instances, and validate recovery objectives.
Connecting to recovery operations…
Instance registry
Consistent backup and verification
API restores are staged and verified. They never overwrite active production files.
Migration journal
Cross-instance recovery transfer
Recovery drill
Operations response
No response yet.
PROJECT / EXTERNAL INTEGRATIONS / v0.38.2
Public API, Webhooks, Embeds, and Research SDK
Operate the stable research API surface, create scoped HTTPS webhook subscriptions, issue signed public-reference embeds, and inspect the dependency-light Python, TypeScript, and browser SDK packages.
Connecting to the public research integration service…
API and SDK discovery
The public catalog and SDK manifest describe stable /v1 routes. Protected operations remain workspace- and scope-governed.
Register an HTTPS webhook
The signing secret is shown only once. Store it outside WordPress and browser code.
Webhook lifecycle and delivery
Outbound network delivery is disabled by default and must be explicitly enabled in the backend environment.
Emit a governed research event
Events contain references, hashes, and policy-approved metadata—not secrets, executable payloads, or restricted dataset bytes.
Create a signed research embed
Embed manifests expire and expose public references only. The browser SDK verifies and renders the returned manifest path.
Integration response
No response yet.
PROJECT / PRODUCT ROUTING / v0.38.1
Typed Cross-Product Research Handoffs
Plan, validate, normalize, and create governed research handoffs through executable product adapters. Every route is contract-bound, target-aware, hash-identified, and delivered through the existing workspace-governed interoperability layer.
Loading typed product adapters…
Typed handoff request
Add both sourceProfileId and targetProfileId to create and seal the handoff in one editor-authorized operation.
Adapter, route, plan, and handoff output
No response yet.
PROJECT / PLATFORM EXCHANGE / v0.38.0
Sustainable Catalyst Research Interoperability Layer
Negotiate typed contracts, create canonical cross-product research handoffs, import compatible envelopes idempotently, and preserve signed delivery receipts without arbitrary callbacks or embedded restricted data.
Connecting to the research interoperability layer…
Workspace and product profile
Compatibility negotiation
Profiles declare supported contract versions and capabilities. The coordinator selects only shared, active contracts.
Canonical research handoff
Import compatible envelope
Imports are canonical-hash verified and idempotent. Duplicate envelopes return the original receipt rather than creating another record.
Receipt and withdrawal
Profiles, negotiations, handoffs, receipts, and provenance
No response yet.
PROJECT / EXPERIMENT CONTROL / v0.33.2
Closed-Loop Simulation and Instrument Campaigns
Connect adaptive campaigns to repeated simulation cycles or signed instrument measurements while preserving operator review, safety interlocks, measurement integrity, and complete campaign/workflow/command provenance.
Closed-loop campaign engine loading…
Loop definition
Loop controls
Safety boundary: Lab emits reviewable command envelopes only. It does not open arbitrary callbacks or directly control physical devices.
Measurement ingestion
Loop, command, measurement, safety, cycle, and provenance records
No response yet.
Persistent Queue Infrastructure
Inspect the durable distributed workload queue, leases, recovery state, and central event history.
Persistent queue loading…
Enqueue and lease
Active leases
| Lease | Worker | Status | Expires |
|---|
Recent queue history
| Time | Entity | Event | ID |
|---|
No response yet.
PROJECT / DISTRIBUTED COMPUTE / v0.31.4
Dispatcher Operations, Dead-Letter Recovery, and Observability
Classify failed workloads, apply bounded retry policies, inspect dead-letter records, replay corrected work, cancel unsafe or obsolete jobs, trace dispatch events, and verify SQLite queue health from one governed operations workspace.
Dispatcher operations loading…
Dead-letter queue
| Queue ID | Method | Project | Failure | Attempts | Dead-lettered | Actions |
|---|
Operator action
Safety: replay and cancellation preserve operator actions and event history. Hard deletion is not exposed.
Operations health, diagnostics, and event timeline
No response yet.
PROJECT / DISTRIBUTED COMPUTE / v0.31.3
Distributed Artifact, Result, and Checkpoint Transport
Move governed inputs, large results, checkpoints, logs, datasets, and provenance records between the coordinator and secure workers with resumable chunks, SHA-256 verification, content-addressed storage, lease-bound access, and retention controls.
Artifact transport loading…
Retained artifacts
| Artifact | Kind | Size | SHA-256 | Owner | Created |
|---|
Transport guarantees
- Content is addressed and deduplicated by SHA-256.
- Chunks are sequential, resumable, and independently verifiable.
- Workers can download only artifacts granted by an active lease.
- Large worker results are externalized from completion receipts.
- Checkpoint and result artifacts retain queue, contract, project, and worker provenance.
- Expired uploads and retained artifacts can be cleaned by policy.
Coordinator transport health and sessions
No response yet.
PROJECT / DISTRIBUTED COMPUTE / v0.31.3
Secure Worker Agent Runtime
Operate pull-based Python workers with one-time enrollment, worker-scoped credentials, local signed-contract verification, governed registered-method execution, automatic lease renewal, provenance-bearing completion receipts, and verified artifact transport.
Secure worker runtime loading…
Registered worker agents
| Worker | Type | State | Mode | Load | Agent |
|---|
Security model
- Enrollment tokens remain on the coordinator and worker host.
- Worker credentials are returned once and stored as coordinator-side SHA-256 digests.
- Dispatch contracts are bound to a worker, signed, and locally verified before execution.
- Only methods registered in the Python Compute Core can run.
- Arbitrary source code, shell commands, callbacks, and executable payloads are rejected.
Local or institutional worker startup
Copy the worker environment example, add the coordinator URL and the two worker secrets on the worker host, validate the configuration, and start the agent.
Coordinator worker-agent health
No response yet.
PROJECT / DISTRIBUTED COMPUTE / v0.31.0
Distributed Compute Dispatcher
Register governed worker capabilities, route registered numerical workloads, issue signed leases, and preserve dispatch records without enabling arbitrary callbacks or arbitrary code.
Compute dispatcher loading…
Worker profile
Workload routing
Worker registry
| Worker | Type | State | Load | Memory | Methods |
|---|
Routing and contract output
Register a worker, route a workload, and issue a governed dispatch contract.
LAB / 0.163.1.2 · SPECIALIST WORKSPACES
Continue from the 4D front door into the specialist research environment.
The 4D visualization and Scientific signals remain visible above. Open one specialist workspace at a time here.
LAB / SPECIALIST WORKSPACES