What Is Workspace Intelligence?
A precise definition of Workspace Intelligence as the evidence-backed understanding layer for software systems.
Workspace Intelligence is the evidence-backed understanding layer for software systems.
Workspai's canonical positioning is Open-Source Workspace Intelligence for Software Systems.
Workspai turns repositories, projects, dependencies, rules, changes, and evidence into shared understanding for developers, CI, IDEs, and AI agents.
Why it exists
Modern software workspaces are not just repositories. They contain projects, services, generated applications, adopted codebases, commands, policies, contracts, CI results, release gates, agent instructions, and operational evidence.
AI tools can read files. Developers can read documentation. CI can run tests. But without a shared layer, each surface builds its own partial truth.
Workspace Intelligence exists so those surfaces can reason from the same evidence-backed operating model.
The definition
Workspace Intelligence answers seven questions:
| Question | Workspace Intelligence answer |
|---|---|
| What exists? | Projects, runtimes, frameworks, commands, policies, reports, and contracts |
| How is it related? | Workspace graph, dependencies, ownership, runtime links, and evidence links |
| What changed? | Snapshots, diffs, changed projects, changed contracts, and changed evidence |
| What is affected? | Impact reports, blast radius, affected commands, and release gates |
| What can be trusted? | Verification status, freshness, policy results, and generated evidence |
| What should agents know? | Agent context, grounding files, skills, and scoped workspace instructions |
| Why does it matter? | Explain and trace outputs for blockers, risks, and release decisions |
What it is not
Workspace Intelligence is not:
- a chat feature,
- a coding agent,
- a generic RAG database,
- a memory store,
- a repository indexer,
- a framework starter,
- a dashboard-only feature.
Those tools can consume Workspace Intelligence. They are not the architecture itself.
The Workspai implementation
Workspai implements Workspace Intelligence as a deterministic CLI and contract layer:
The CLI writes machine-readable reports and agent-facing artifacts from the same workspace evidence. That keeps public documentation, CI, IDE surfaces, and AI grounding aligned with the actual workspace state.
Evidence is the authority
Documentation explains the architecture. Evidence decides what is true for a workspace.
That is the core rule:
Evidence is the source of understanding; documentation is an output, not the authority.
In Workspai, claims should come from contracts, reports, verification gates, workspace metadata, or generated artifacts. If a feature is planned but not implemented, it belongs in an RFC or roadmap, not in an implemented claim.
Where to go next
- Concepts for the vocabulary.
- Context vs Understanding for the difference between more input and better system meaning.
- Repository Intelligence vs Workspace Intelligence for the product boundary.
- Workspace Model for the canonical representation.
- Evidence for the trust model.