Knowledge & review

Knowledge base

What it stores, how analyze ingests it, and why it cuts re-writing.

KnowledgeArtifact stores reusable code typed as function, helper, service, component, widget, module, interface or pattern — each with name, language, framework, summary, file_refs, structure, source_meta and a pgvector embedding. ProjectContextFile stores versioned context (codebase / workflow / notes); RepoContext stores framework, tech stack, architecture summary and directory map.

Knowledge ingestion
  1. 1

    Run analyze

    maestro analyze

    Architect permission required.

  2. 2

    Analysis run

    POST /projects/{id}/analysis

    Creates an AgentRun with purpose repo_analysis.

  3. 3

    Context emitted

    RepoContext + versioned ProjectContextFiles (prior version snapshotted on change).

  4. 4

    Artifacts indexed

    KnowledgeArtifacts written with embeddings; ContextFileUpdatedBroadcast fired.

  5. 5

    Reuse before writing

    Future tasks/agents query the catalog before coding.

  6. Loops back until the verdict is approved

Why it reduces re-writing code

Agents receive prior architecture, conventions and a catalog of existing functions, services and components — so they extend what exists instead of re-implementing it. Fewer tokens, fewer duplicate implementations, more consistent code.

Search

GET /api/v1/projects/{uuid}/knowledge and the Studio /knowledge browser. Filters by type, language, framework; free-text q runs an ILIKE across name, summary and framework, scoped to the user’s workspace/project.