Openkb

What It Is

OpenKB is an open-source Python CLI by VectifyAI that compiles a folder of raw markdown into an LLM-synthesized wiki. Each source document gets a summary page; cross-cutting themes get auto-generated concept pages; queries can be saved as exploration artifacts; and the resulting graph can be linted for orphans, contradictions, and gaps. It is bring-your-own-LLM (any OpenRouter-compatible model, including free Llama 3.3 70B).

For SIE-adjacent work, OpenKB is most useful as a reference implementation — a smaller, narrower tool that solves a sliver of what SIE solves, but does a few things SIE currently doesn’t. This doc captures those patterns.

Architecture & Concepts

Directory Taxonomy

OpenKB enforces a strict split between source-derived and synthesis content:

Compile Workflow

1. openkb add <doc.md> — LLM reads the document, writes a summary page, and updates/creates concept pages where the doc relates to existing themes. Cross-references emerge as Wikilinks.
2. openkb list / openkb status — inspect indexed content.
3. openkb query "..." — natural-language Q&A against the synthesized wiki.
4. openkb query "..." --save — persist the answer as an explorations/ page.
5. openkb lint — health check across the wiki, output to reports/.
6. Adding a new doc later updates only the affected concept pages (incremental, not full rebuild).

Lint Checks

The lint command flags:
– Orphans — pages with no inbound or outbound Wikilinks.
– Contradictions — claims across pages that disagree.
– Gaps — concept pages referenced but not yet written.
– Stale links — Wikilinks whose target was renamed or removed.

Programmatic Graph Analysis

OpenKB exposes the wiki structure as plain markdown, so a small Python script can compute:
– Inbound link counts per page (hub identification)
– Cross-reference adjacency (which pages link to which)
– Page size distribution and link density

SIE Parallels & Adaptable Patterns

OpenKB and SIE solve overlapping problems with different scope. SIE is the production engine for WordPress-bound publishing across multiple sites; OpenKB is a local research wiki. The following OpenKB patterns are worth porting into SIE:

What SIE Already Does Better

• WordPress-bound publishing pipeline with REST sync.
• Multi-site mother/child architecture (bernardhats, toppizza, etc.).
• Schema discipline across SI fields (post 2026-04-25 alignment).
• FAQ-profile taxonomy and Rank Math integration.
• Domain-specific verification (e.g. WB anecdote checks against primary sources).

OpenKB has none of these — it stops at the local wiki layer.

Limitations vs. SIE

• No publishing pipeline. Output is local markdown only.
• No schema enforcement. Frontmatter is freeform; no SI fields, no FAQ profiles.
• Lint heuristics are LLM-judged, so quality varies with model strength.
• No multi-site / mother-child concept. One wiki per directory.
• No verification layer for factual claims against external sources.

Pricing

• Tool itself: Free, open source, MIT license.
• LLM costs: BYO OpenRouter (or compatible) API key. Works with free models like meta-llama/llama-3.3-70b-instruct:free for zero-cost experimentation.

Expert Notes

The highest-leverage import from OpenKB into SIE is the lint command. Building a SIE-side equivalent — even a minimum-viable version that just flags orphan pages and missing Wikilink targets — would have immediate payoff on the active cinema and WB-POI cluster expansion work. The explorations/ pattern is the second priority: a synthesis-question audit trail compounds in value.

OpenKB itself is not worth adopting as tooling — SIE already covers its scope and integrates with the WordPress publishing layer. Treat it as a design reference, not a tool to install.

Direct Link: https://github.com/VectifyAI/OpenKB