wikify

Researcher-style iterative wiki builder. Editor orchestrator dispatches explorer subagents that walk the corpus via named recursive patterns (P1-P5), gather evidence into notebook dossiers, and write pages when a composite maturity score crosses the gate. A DATA wave harvests verifiable numbers/tables into a claim store and consolidates them into evolving kind=data artifact tables. Coverage of the corpus chunk set is the primary objective. Re-entrant on the same bundle when new corpus material arrives.

Recursive exploration pattern library for wikify. Five named, depth-bounded procedures (P1 hub-anchor, P2 citation-walk, P3 semantic-boundary, P4 exact-term sweep, P5 gap-explorer) that compose existing corpus_find / corpus_traverse / corpus_citation_walk / wiki_find MCP primitives and append evidence into per-slug notebooks. Editor invokes one pattern per Task with a slug-disjoint target list.

Shared Wikify reference material for Claude Code skills. Use when working on Wikify skills, workflows, prompts, bundle state, writing rules, schemas, citation grounding, CLI grammar, exploration patterns, or MCP setup.

Produce a Wikify WriteResponse from supplied page context and evidence. Use when a workflow has selected a concept/page and needs encyclopedic article, person-page, comparison, or refinement prose grounded in evidence. Does not commit pages or decide readiness.

Consolidator that turns the bundle claim store into evolving data-artifact tables. Picks a dense subject-by-property theme, writes a durable view spec, and commits a kind=data wiki page that renders as an HTML table with per-cell references and re-derives from the spec as more data lands.

Extractor that harvests verifiable numeric/factual data points from corpus chunks, tables, and figure captions into the bundle claim store. Runs as a dedicated harvest pass over table-assets plus number-dense chunks, or piggybacks on chunks an explorer just read. Every point carries a verbatim grounding quote; the CLI gate rejects numbers it cannot locate in the source.

Canonical evidence-gathering skill. A sonnet supervisor plans one shared query plan for all sibling slugs in a cluster, fans out cheap haiku judges that read chunks and emit per-chunk routing+score+quote, then commits one evidence ledger per slug. Handles clusters of any size (including singletons).

Parse a directory of documents (PDF / DOCX / PPTX / HTML) into a queryable Wikify corpus. Use when the user has local files to turn into a corpus, or to finish an arxiv harvest. Wraps `wikify corpus build` and owns parser-backend choice (docling default, lite), the --out convention, partial-failure policy, and post-build health checks.

Acquire arXiv papers on a topic and stage them for `corpus build`. Use when the user wants to download, harvest, fetch, or bulk-collect arXiv papers about a subject, field, or category, or seed a Wikify corpus from arXiv. Composes `arxiv scout` (discover categories), `arxiv identify` (harvest metadata), `arxiv download` (PDFs), then `corpus build`, owning category selection, coverage threshold, resumability, and partial-failure policy.

Query workflow for answering from the committed Wikify wiki, falling back to corpus search when needed, and recording bundle feedback for refinement. Use when answer sufficiency, corpus fallback, and feedback policies are supplied.

Build a wiki end to end with the simplified conventional-RAG pipeline (the simple path, not the primary researcher-style `wikify` builder): extractor-driven concept suggestions, optional author/person seeding, evidence gathering, writing, validation, commit, render, eval, inspection, and iteration without hand-curated concept lists.

Explain and use mechanical Wikify bundle operations: run lifecycle, work concepts, evidence, feedback, claims, draft build/check, validation, commit gates, projections, render, eval, events, locks, and failure codes. Use only for state inspection and mutation mechanics; strategy belongs in workflow skills.

Build a validated navigation hierarchy for a committed Wikify bundle. Use after batches of pages are committed and before render so articles are grouped into reader- and agent-friendly topic sections.

Refinement workflow that consolidates bundle inbox feedback, finds concepts marked needs_refine, gathers additional evidence when required, rewrites pages, validates, and commits replacements. Use when refinement threshold, batch, and retry policies are supplied.

Explain and use the Wikify corpus CLI as the read/search surface over the corpus fluent API. Use when probing corpus documents, chunks, authors, citations, figures, equations, sampling diverse entry points, semantic search, text search, paper-level ranking by citation count, or recursive graph traversals. This skill is read-only and does not decide an exploration strategy.

Use the committed-wiki MCP tools (or bash CLI as fallback) to search pages, traverse wiki relations, and bridge back to the corpus.