wiki

Manages an Obsidian wiki vault through eight modes covering ingest, query, lint, weave, report, init, extract, and format.

Updated Aug 2, 2026
One-click install
npx skills add https://github.com/phm-aguiar/test_documentacao --skill wiki-phm-aguiar
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: wiki
Source: https://github.com/phm-aguiar/test_documentacao/tree/main/.claude/skills/wiki
Command: npx skills add https://github.com/phm-aguiar/test_documentacao --skill wiki-phm-aguiar

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires sentence-transformers, rank_bm25, numpy, scikit-learn, openai, anthropic.

What problem does it solve? Maintaining a personal knowledge base in Obsidian involves many fragmented operations—ingesting sources, answering queries, auditing link health, deduplicating pages, and generating reports. This Skill consolidates roughly 30 atomic wiki operations into one toolkit with 8 modes, so a single entry point handles the full knowledge lifecycle. ## Core Features & Use Cases - Ingest & Query: Incrementally ingest Markdown, PDF, JSON/CSV, images, and URLs into a vault with manifest tracking, then query via tiered retrieval (index grep, optional QMD vector search, section grep, full read) including multi-hop relationship traversal. - Lint & Weave: Run 13 health checks (orphans, broken links, stale content, tag cohesion) with optional consolidation, plus cross-linking, dedup merging, synthesis page generation, and tag taxonomy normalization. - Report & Init: Generate status deltas, dashboards (Obsidian Bases/Dataview), digests, and graph exports (JSON, GraphML, Cypher, HTML), or bootstrap a new vault with directory structure, config, and manifest files. - Use Case: After dropping research notes into your vault, run the ingest mode to create structured pages with frontmatter and relationships, then weave cross-links and run lint to audit the resulting graph. ## Quick Start Ask the assistant to run the wiki skill in ingest mode to add a new source document to your Obsidian vault.

Frequently Asked Questions about wiki

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I ingest documents into an Obsidian vault automatically?▼

Use the ingest mode, which reads sources (Markdown, PDF, JSON/CSV, images, URLs via defuddle), extracts concepts and relationships, and writes pages with frontmatter. Incremental append mode uses a SHA-256 manifest so only new or modified sources are processed.

How to find broken links and orphan pages in Obsidian?▼

Run the lint mode, which performs 13 checks including broken wikilinks, orphaned pages with zero incoming links, missing frontmatter, and stale content. A fast sub-mode runs the three core checks in under three seconds via a script.

Does the wiki query mode support semantic search?▼

Yes, query mode optionally uses QMD for lexical plus vector search when the QMD_WIKI_COLLECTION variable is configured. The companion experiential_memory module also provides cosine similarity and hybrid BM25 search over sentence-transformers embeddings stored in SQLite.

Can I merge duplicate notes in my Obsidian wiki?▼

Use weave mode with the dedup sub-mode, which computes similarity via token overlap, edit distance, and alias matching with a 0.75 threshold. It audits first, then merges by picking a canonical page, creating a redirect stub, and rewriting wikilinks.

What Python dependencies does the wiki embedding index require?▼

The experiential_memory module requires sentence-transformers (all-MiniLM-L6-v2 model) and rank_bm25 for retrieval. numpy and scikit-learn are optional for KMeans clustering, and openai or anthropic SDKs enable LLM-based distillation of canonical chunks.

Why is my wiki query returning stale or archived pages?▼

Pages carry lifecycle fields in frontmatter, and query results annotate stale or archived entries explicitly. Run lint mode to detect stale content and use the consolidate sub-mode to advance lifecycle states or add stale callouts.