KairoECS Documentation

KairoECS is a Rust-first simulation engine for deterministic event scheduling,

ECS-style state, DES, ABM, Arrow telemetry, and polyglot bindings.

Public project links:

• Repository: github.com/edithatogo/kairos
• CI status: GitHub Actions
• Conductor status: track index and narrative status

Current maturity: pre-release. Core scheduler/state/conformance foundations are

implemented; DES/ABM flow APIs and the FFI bridge are in review; registry

publication and stable API promises remain gated by release, compatibility,

security, and packaging tracks.

Start Here

• Documentation overview
• Install and local workflow
• Conductor track status
• Current narrative status
• Benchmark overview
• Implemented crate inventory
• Language binding inventory
• Documentation examples
• Community adoption
• Community playground
• Model zoo inventory
• Starter kits
• Scenario run and replay
• Experiment runner CLI
• Citation and archival
• Developer workflow
• Trustworthy simulation
• Factory bottleneck VVUQ note
• Visualization guide

Tutorials and Examples

• Tutorial index: learning paths, Rust getting started, Python getting started, Wasm and TypeScript getting started, and model-building.
• Binding quick lessons: R getting started, Julia getting started, C# getting started, and Go getting started.
• Example surfaces: scenario run and replay, headless snapshot playground, and documentation examples.
• Jupyter notebooks and figures: notebook and image assets are tracked by the examples and education tracks; this site manifest keeps those entry points discoverable without vendoring large generated media into the static shell.

Language Quickstarts

• Python binding
• R binding
• Julia binding
• TypeScript/Wasm binding
• C# binding
• Go binding

Governance and Release

• API review template
• API review
• Compatibility governance
• Interoperability standards
• Release governance
• Release checklist
• Release compatibility
• Supply-chain verification
• Changelog policy
• Maintenance handoff

Specialized Systems

• Cloud HPC policy
• AWS Batch
• Azure Batch
• GCP Batch
• Slurm
• Distributed deployment
• Transport trait
• Telemetry aggregation
• Entity migration protocol
• FMI import guide
• FMI export guide
• GPU compute architecture
• GPU benchmark results
• ML architecture
• PDES logical-process trait
• Streaming architecture

Trust and Quality

• Benchmark policy
• Reproduce comparison
• Research reproducibility
• VVUQ note
• Replay and seeds
• Scenario evidence
• Trace format
• CLI reference
• Community governance
• Community roadmap

R2 Implementation Map

• Core and shared contracts: kairo-ecs-types, kairo-ecs-core, kairo-ecs-state, kairo-ecs-rng, kairo-ecs-des, kairo-ecs-abm, kairo-ecs-arrow, and kairo-ecs-ffi.
• Binding bridges and facades: kairo-ecs-uniffi, kairo-ecs-diplomat, kairo-ecs-wasm, plus preview language packages for Python, R, Julia, TypeScript/Wasm, C#, and Go.
• Optional and advanced tracks now represented in docs: visualization, GPU/WebGPU, PDES, distributed MPI/gRPC, streaming, ML inference, FMI digital twin, cloud/HPC, and time-travel debugging.
• Native binding status remains explicit: preview packages may expose pure-language scheduler facades while reporting native FFI as not configured until Track 02 publishes stable local artifacts.
• Maturity language is staged: alpha means evidence-backed but limited, beta means wider release gates are active, and stable remains blocked until compatibility and release-candidate gates pass.

Local workflow

• npm --prefix website run check:all
• just docs-build
• npm --prefix website run check:links
• npm --prefix website run check:quality
• just docs-dev
• just check-docs
• just validate-conductor
• just validate-tracks
• just validate-track-docs
• just validate-conformance
• just dev-validate

Current docs tree

• docs/adr/ for architecture decisions, naming, and release staging.
• docs/api/api-review-template.md and docs/design/api-review.md for API review intake.
• docs/benchmarks/ for benchmark overview, policy, and reproducibility guidance.
• docs/cloud-hpc/ for AWS Batch, Azure Batch, GCP Batch, Slurm, and spot-checkpoint policy notes.
• docs/community/ for contributor onboarding, governance, adoption, model-zoo guidance, roadmap notes, and the playground.
• docs/debugging/ for trace format and interactive debugging CLI references.
• docs/cli/ for the scenario runner CLI quickstart and command reference.
• docs/distributed/ for deployment, transport, telemetry aggregation, and entity migration notes.
• docs/fmi-digital-twin/ for FMI import/export, AAS mapping, and deployment model notes.
• docs/gpu-compute/ for GPU/WebGPU architecture, hardware requirements, benchmark results, and maintainer notes.
• docs/interoperability/standards-review.md for interoperability standards review.
• docs/ml/ for ML integration architecture, model versioning, and surrogate authoring.
• docs/pdes/ for PDES protocols, logical-process traits, GVT, and benchmark results.
• docs/release/ for release, compatibility, maintenance, changelog, and supply-chain checklists.
• docs/research/ for citation and reproducibility guidance.
• docs/scenarios/ for committed run/replay scenario notes.
• docs/starter-kits/ for starter-kit discovery and maturity guidance.
• docs/streaming/ for real-time processing architecture, stream schema, and broker setup.
• docs/trustworthy-simulation/ for replay, seeds, and uncertainty notes.
• docs/tutorials/ for learning paths, notebooks, and language quickstarts.
• docs/validation/ for fixture-backed VVUQ notes.

Documentation Quality Gate

• website/docs-link-manifest.json lists the source docs and binding pages the site must keep reachable.
• npm --prefix website run check:links verifies required paths, navigation targets, and local Markdown links without writing generated output.
• npm --prefix website run build renders website/build/index.html, source-backed HTML pages for each Markdown navigation target, website/build/docs-index.json, website/build/sitemap.xml, and website/build/robots.txt; the build output is ignored by git.
• npm --prefix website run check:quality verifies build artifacts, docs concepts, navigation coverage, generated source pages, and the static output size budget.
• The generated docs index is intentionally offline-first and dependency-light, so CI and local contributors do not need network access after checkout.

Contributor commands

• cd website && npm ci && npm run build
• cd website && npm run check:links
• cd website && npm run check:quality
• cd website && npm start
• just docs-dev

Site owner

Track 14 owns the public docs surface, while Track 27 owns the contributor workflow commands that keep the site buildable.