Repository Intelligence vs Workspace Intelligence
The difference between understanding a codebase and understanding a software system.
Repository Intelligence and Workspace Intelligence are related, but they are not the same layer.
Repository Intelligence helps tools understand a codebase.
Workspace Intelligence helps tools understand the software system around that codebase.
The distinction
A repository is usually a source boundary.
A workspace is an operating boundary.
That operating boundary can include:
- multiple projects,
- generated and adopted applications,
- runtime commands,
- policies,
- contracts,
- CI evidence,
- release gates,
- ownership,
- agent grounding,
- historical model snapshots,
- impact and verification reports.
Repository Intelligence
Repository Intelligence usually focuses on what exists inside a repository.
It can answer:
- Which files exist?
- Which symbols exist?
- Which files import each other?
- Which functions call each other?
- Which areas changed recently?
- Which owner or team touched this code?
- Which code paths look important?
Typical outputs:
- code graph,
- symbol graph,
- file summaries,
- search index,
- ownership hints,
- repository health signals.
This is valuable. Workspai should learn from it. But it is not the whole system.
Workspace Intelligence
Workspace Intelligence starts from a different question:
What does this change mean for the software system?It can connect repository facts to:
- workspace model,
- project registry,
- runtime commands,
- dependency graph,
- policies,
- contracts,
- doctor evidence,
- readiness gates,
- impact analysis,
- verification,
- agent context,
- generated grounding artifacts.
Typical outputs:
- workspace model,
- workspace graph,
- workspace context,
- model diff,
- impact report,
- verify report,
- agent customization pack,
- explain or trace narrative.
Comparison
| Question | Repository Intelligence | Workspace Intelligence |
|---|---|---|
| What is the boundary? | Repository | Workspace / software system |
| What is the main object? | Files, symbols, imports | Projects, runtime, policies, evidence, agents |
| What graph does it build? | Code graph | Software system graph |
| What does it optimize for? | Codebase navigation and retrieval | Shared operational understanding |
| What does it help agents read? | Source code and repository summaries | Scoped, evidence-backed workspace context |
| What does it help CI decide? | Usually limited | Impact, verify, release readiness |
| What is the trust model? | Often inferred from code/index | Evidence, freshness, contracts, gates |
Why this matters for AI agents
AI agents do not only need to know which file changed.
They need to know:
- which project is affected,
- which dependent projects may be stale,
- which commands verify the change,
- which release gates apply,
- which context should be refreshed,
- which agent-facing artifacts need to be synced,
- which facts are verified, observed, inferred, stale, or unknown.
Repository Intelligence can help with the file and code layer. Workspace Intelligence connects that layer to the rest of the system.
Workspai boundary
Workspai does not replace Repository Intelligence.
It adds the Workspace Intelligence layer above it:
That is why the product category is not:
better repository searchIt is:
Workspace Intelligence for software systemsPractical rule
Use this distinction:
Repository Intelligence explains codebases.
Workspace Intelligence explains software systems.If the question is "where is this function used?", Repository Intelligence may be enough.
If the question is "what does this change affect, what evidence proves it, and what should agents trust?", you need Workspace Intelligence.