RapidKit Labs is the home of Workspai, RapidKit Core, and the contracts that connect developer workflows across CLI, IDE, CI, and AI.
| Product | Role | Source |
|---|---|---|
| Workspai CLI | Open-Source Workspace Intelligence for Software Systems | rapidkitlabs/workspai |
| Workspai for VS Code | Visual workspace operations, evidence, repair, and AI workflows | rapidkit-vscode |
| RapidKit Core | Python engine for generators, modules, and doctor workflows | rapidkit-core |
| Reference Workspaces | Reproducible examples and adoption patterns | rapidkit-examples |
Workspai turns projects, policies, evidence, and changes into one shared model that developers, CI, IDEs, and AI agents can use.
flowchart TB
INPUTS["Software system inputs<br/>Projects · Policies · Contracts · Changes · Runtime"]
INTELLIGENCE["Workspace Intelligence<br/>Model · Graph · Context · Impact · Verify · Explain"]
OUTPUTS["Evidence-backed outputs<br/>Reports · AGENTS.md · Skills · Freshness · Release gates"]
CONSUMERS["Developers · CI · IDEs · AI agents · MCP clients"]
INPUTS --> INTELLIGENCE --> OUTPUTS --> CONSUMERS
CONSUMERS -. outcomes and fresh evidence .-> INPUTS
classDef core fill:#0f172a,color:#f8fafc,stroke:#38bdf8,stroke-width:2px;
class INTELLIGENCE core;
Start with software you already have:
npx workspai adopt ../existing-project --json
cd ~/.workspai/workspaces/workspai
npx workspai workspace model --json --write
npx workspai workspace context --for-agent --json --write
npx workspai workspace verify --strict --jsonInstall globally when preferred:
npm install -g workspai| Domain | Responsibility |
|---|---|
| getrapidkit.com | RapidKit Labs brand and product ecosystem |
| workspai.dev | Workspace Intelligence knowledge portal |
| workspai.com | Workspai product and online experience |
This is the ecosystem governance hub. It contains no runnable product code. Use it for cross-repository architecture, roadmap, contribution guidance, and release coordination.
Implementation issues belong in the repository that owns the behavior.
Evidence before confidence. Architecture before feature count.