Knowledge Connector

Problem

When notes and documents are scattered across files, repeated concepts are hard to trace, and users often stop at “stored” rather than “understood.” Simple concept extraction produces isolated entities without source coverage, relationship strength, or next actions, which easily becomes another generic knowledge-graph utility.

How It Works

knowledge-connector organizes import, retrieval, visualization, and next-step guidance:
- Import: use kc import-docs to build a graph from files or a directory, and kc import-wizard for a preview-first onboarding flow while preserving source titles and paths.
- Cross-document search: use kc search / kc query to find where a concept appears across documents and which concepts connect multiple sources.
- Visualization: use kc visualize for full graph export and kc map for a concept-centered actionable subgraph.
- Output shape: prefer What Matched, Why It Matters, and Next Step, explaining matching concepts, source coverage, the meaningful pattern, and actions such as importing more material, inspecting weak relationships, or exporting the graph for review.

Boundaries

It fits a growing notes corpus, repeated cross-file concepts, and moving from storage to understanding. It is weaker when it behaves only as a raw extractor without import flow, source-aware search, and next-step guidance.