Workspai.dev

Workspai.dev

The official learning path for Workspace Intelligence.

Workspai.dev is the public knowledge portal for Workspace Intelligence: the evidence-backed architecture layer that helps humans, CI, IDEs, and AI agents understand the same software system.

Workspai turns repositories, projects, dependencies, rules, changes, and evidence into shared understanding for developers, CI, IDEs, and AI agents.

This site teaches the architecture. It is not the cloud app, not a dashboard, and not a repository upload surface.

Learning path

1. Concepts

Understand why Workspace Intelligence is different from RAG, memory, repository indexing, and chat.

2. Software System Understanding

Learn the structure, relationships, intent, evolution, trust, and reasoning layers an AI needs.

3. Mental Models

Reason about shared system models, evidence-backed context, workflow boundaries, and agent authority.

4. Architecture

Study the input layer, Workspace Intelligence core, and consumers.

5. Workspace Model

See how projects, runtimes, commands, policies, contracts, and evidence become one model.

6. Workspace Graph

Separate code graphs from a graph of the software system: structure, runtime, ownership, change, and evidence.

7. Evidence

Learn how reports, contracts, gates, freshness, and verification keep claims trustworthy.

8. Agent Grounding

Understand how AGENTS.md, context packs, skills, IDE files, and MCP access align AI tools.

9. Create and Adopt

Learn how existing projects enter through adopt/import and which supported kits can start new projects.

10. Verified Engineering Goals

Turn release, dependency-security, and coverage requests into durable definitions of done.

11. Verified AI Repair

Follow causal findings through one bounded transaction, verification, rollback, or a durable decision.

12. Agent Skills

See how runtime-aware Skills and canonical bootstrap keep model hosts aligned without duplicating authority.

13. Agent Entry and Bootstrap

Resolve any adopted project to its governing workspace and validate the bounded agent starting point.

14. Analysis, Readiness, and Verification

Separate findings, release preparation, and the final governed verdict.

15. Change Intelligence

Connect snapshots and diffs to blast radius, traces, and proof-backed explanations.

16. Release Governance

Understand contract gates, strict policy, the broader pipeline, and release authority.

17. Observation and Evaluation

Observe changes, record structured outcomes, and evaluate agent value without storing conversations.

18. Machine Integrations and Portability

Learn the MCP, export, sharing, hydration, archive, and restore boundaries.

19. Modules and Dependencies

Operate runtime-aware modules and dependency repair with checkpoint and verification boundaries.

20. Capability Coverage

Audit the enforced map from important capabilities to learning guides, commands, and contracts.

21. Contract Catalog

Inspect every versioned contract, field, artifact, producer, and consumer boundary.

22. Command Catalog

Follow every CLI command to its canonical execution, artifacts, and governing contracts.

Extended CLI domains

Publications

What this portal should help you answer

  • What is Workspace Intelligence?
  • What is the difference between a repository graph and a workspace graph?
  • Which facts are verified, observed, inferred, stale, or unknown?
  • How does a project enter the workspace: native, official, or existing?
  • What context should an AI agent receive before it acts?
  • Which commands create model, context, impact, verification, and release evidence?
  • Which public claims are allowed by the shipped contracts?

Boundaries

Workspai.dev stays focused on public education: concepts, architecture, mental models, contracts, commands, guides, essays, research, RFCs, glossary, and release notes.

Interactive product experiences, hosted workspace operations, live repository analysis, account workflows, and commercial surfaces belong on workspai.com.

Chistiq company and portfolio pages belong on chistiq.com. RapidKit product pages belong on getrapidkit.com.

On this page