Decision intelligence that explores multiple futures before you commit.

Chronos is a temporal compute platform: it plans work, simulates possible strategies, evaluates trade-offs, and ranks the strongest path — for people and autonomous agents.

🌐 Live: chronoslab.space

Link previews (X, Slack, iMessage, LinkedIn) use the Open Graph card at
https://chronoslab.space/og-image.png (public/og-image.png).

Most AI returns a single answer. Chronos returns a decision:

Goal → Gather context → Generate futures → Evaluate trade-offs
    → Rank outcomes → Recommend the best path

Use it when the cost of a wrong path is high — product launches, capital allocation, research strategy, or agent planning that must think before it acts.

# Clone
git clone https://github.com/Chronos-Lab-Space/Chronos.git
cd Chronos

# Install
npm install

# Configure
cp .env.example .env
# Set VITE_SUPABASE_URL and VITE_SUPABASE_ANON_KEY

# Apply database schema (Supabase SQL editor)
# Prefer migrations under supabase/migrations/
# or bootstrap with supabase/schema.sql

# Dev
npm run dev

# Unit tests
npm run test:unit

# Production build (includes GH Pages 404 fallback)
npm run build

Open http://localhost:5173.

See .env.example.

src/
├── domain/           # Pure models: Chronos engine, workspace types, gates
├── application/      # Use cases: planner, simulation engine, workspace service
├── infrastructure/   # Supabase, auth, local store, caches, repositories
├── presentation/     # React app, marketing pages, workspace UI
│   ├── components/   # Site shell, docs, FAQ, changelog
│   └── features/     # Workspace, knowledge, simulation, memory, planner
├── main.tsx
└── index.css
supabase/
├── migrations/       # Schema evolution (workspaces, sims, versioning, …)
├── functions/grok/   # Edge Function → api.x.ai
└── schema.sql        # Bootstrap reference (keep aligned with migrations)

Architecture rules: ARCHITECTURE.md
Performance notes: PERFORMANCE.md
Testing strategy: TESTING.md

Legacy /dashboard redirects to /workspace.

Sign in → Create workspace → Set goal → Upload knowledge
       → Run simulation → Review report & timeline → Re-run / memory

Persistence is local-first with cloud dual-write:

• localStorage for instant resume
• Supabase for durable multi-session memory when authenticated
• Load merges remote + local simulation history and backfills empty cloud from local

The home live demo and /simulate share publicStartupSimulator:

1. Decompose the objective into a task graph
2. Simulate ranked go-to-market paths
3. Collapse to best path + alternatives (ARR, probability, roadmap)

Deterministic for a given prompt (cacheable).

npm run dev        # Vite dev server
npm run build      # Production bundle + 404.html for SPA hosts
npm run preview    # Preview production build
npm run test:unit  # Vitest
npm run test:e2e   # Playwright (install browsers first)

npx playwright install chromium
npm run test:e2e

Timeline → Branch → Evaluate → Prune → Collapse → Memory

Chronos executes tasks and capabilities, not fixed agent personas. Planners build dependency graphs; the runtime forks futures, scores them, and keeps lineage for audit and re-runs.

Agent OS sketch:

Planner → Task Graph → Scheduler → Execution
        → Memory → Evaluation → Timeline ranking

This repository is the public product surface for Chronos Lab (site, workspace UI, docs, simulator). Bug reports and suggestions are welcome via GitHub Issues.

MIT © 2026 Chronos Lab. See LICENSE.

• Website: chronoslab.space
• Docs: chronoslab.space/docs
• X: @chronoslabspace
• Telegram: join group
• GitHub: Chronos-Lab-Space/Chronos