obsidian

Plugin obsidian from marketplace jylee-claude-marketplace in JuyeongYi/jylee_claude_marketplace. Commands read from the repo's manifest — 1 month ago.

Skills (44)

• obsidian:aliases

  Query Obsidian wikilink aliases — list aliases for a specific file or the entire vault. Useful for discovering alternate names a note answers to.

• obsidian:base

  Manage Obsidian Base files (database) — list bases, run queries, list and create views over the vault as a structured dataset.

• obsidian:bookmarks

  Manage Obsidian bookmarks — list saved bookmarks (notes, folders, searches) and inspect details.

• obsidian:bulk-scan

  Batch-collect vault metadata using obsidian list commands and qmd multi_get. Avoids N individual CLI calls for large-scale operations across many files.

• obsidian:consistency-check

  Detect contradictions between vault notes on overlapping topics. Cluster notes by tag/folder/semantics, compare factual claims, and report conflicts with resolution suggestions.

• obsidian:daily

  Read and write Obsidian daily notes — open today's note, append/prepend content, resolve the daily-note path.

• obsidian:diff

  Compare local and Obsidian Sync versions of a file — list available versions and diff between any two to see what changed.

• obsidian:file-mgmt

  Move, rename, and delete vault files. Destructive — confirm with user before executing each operation.

• obsidian:graph

  Analyze the vault's link graph — incoming links (backlinks), outgoing links, and graph anomalies (orphans, deadends, unresolved). Core data source for vault-healer.

• obsidian:history

  Obsidian file history — list version history, read past versions, and restore a file to an earlier state. Useful for undo, accidental delete recovery, and version comparison.

• obsidian:inbox-convert

  Convert non-markdown files in Inbox/ to markdown using markitdown. First step of the inbox-ingest pipeline. Also processes Inbox/yt.md — a URL-per-line list — into per-URL markdown files.

• obsidian:inbox-extract

  Extract chapter/topic-level knowledge from large converted documents (books, long PDFs, reports) and match each chunk against existing vault notes for merging. Bridges inbox-convert and vault-ingest.

• obsidian:inbox-extractor

  Decompose large converted markdown into topic-based chunks, match each against vault knowledge via qmd, and emit per-chunk files into Inbox/.extracted/ for vault-ingester.

• obsidian:inbox-ingest

  Full Inbox processing pipeline. Convert non-markdown files, analyze and merge into vault, update search index, add related links, and extract glossary terms.

• obsidian:inbox-link-weave

  Suggest and apply related-note links for notes that were just created or merged by vault-ingest. Targets only the recently-touched set, not a full vault scan.

• obsidian:inbox-vocab-update

  Extract technical terms from recently ingested notes and create stub glossary entries under the vault's glossary folder for later expansion.

• obsidian:knowledge-enrich

  Enrich stub and draft notes with content from vault context, source URLs, and web research. Drives notes from stub to draft to complete.

• markitdown

  Convert various file formats to markdown using the markitdown CLI. Supports PDF, HTML, DOCX, PPTX, XLSX, CSV, images, audio, EPUB, Jupyter notebooks, Outlook MSG, ZIP, and plain text.

• obsidian:note-scorer

  Score the quality of a vault note (0–100) based on word count, heading structure, link density, source presence, and status accuracy. Used by vault-healer, knowledge-enricher, and vault-stub-reporter; also callable directly.

• obsidian:note-splitter

  Split an oversized vault note into focused child notes. Analyze headings, pick a split strategy, create child notes, and rewire wikilinks. Called by vault-healer or directly.

• obsidian:notes

  Read, create, append to, and prepend content of vault notes — primary file-content CRUD for any markdown note.

• obsidian:obsidian-search

  Search the vault and summarize results. Find related notes by query, read their content, and provide an answer.

• obsidian:properties

  Inspect and modify note frontmatter properties — set/read/remove a single property on one file, or list properties used across the whole vault.

• qmd-collection

  Manage qmd collections — add/remove/list indexed folders, attach context meta descriptions, and browse indexed files with ls.

• qmd-get

  Retrieve specific documents from qmd index by path or docid. Use for post-search retrieval, reading specific files, and batch retrieval.

• qmd-index

  Maintain qmd index — check status, reindex, refresh vector embeddings, clean up. Run embed in background as it requires model loading.

• qmd-refresh

  Run qmd update and embed to refresh search index

• qmd-search

  Search local markdown knowledge base using qmd CLI. Provides MCP-based hybrid semantic search (lex+vec+rerank) and BM25 keyword search.

• obsidian:recents

  List recently opened files in the Obsidian workspace, optionally limited to a count.

• obsidian:search

  Raw text search inside vault files — exact-string and substring matching with optional folder filter and per-match line context.

• obsidian:structure

  Explore vault file/folder layout — list files (filtered by folder/extension), enumerate folders, surface vault metadata, and outline a note's heading structure.

• obsidian:synthesize

  Read a subtitle file (.srt) and synthesize a Korean structured summary into a companion markdown note. Replaces raw-transcript embedding with a 09 Geometry Selection-style note — 핵심 포인트 + 실습 흐름 + 관련 개념. Used as the final step after transcribe acquires subtitles.

• obsidian:tags

  Query and analyze vault tags — list all tags, count occurrences, drill into a single tag, and inspect a file's tag set. Used by vault-classifier to learn the existing taxonomy.

• obsidian:tasks

  Manage vault checkbox tasks — list todo/done items across the vault or one file, and toggle status of individual tasks.

• obsidian:transcribe

  Transcribe audio or video (local file or URL) to an SRT sidecar using ffmpeg 8.0's native Whisper filter. Produces a subtitle file only — it never touches markdown notes. Pair with the `obsidian:synthesize` skill to turn the SRT into a structured Korean summary note.

• obsidian:vault-classify

  Auto-classify untagged notes. Learn existing tag/property patterns, suggest tags/properties for unclassified notes, and apply upon approval.

• obsidian:vault-git

  Git operations scoped to the vault — incremental change tracking, auto-commit after agent runs, and listing files changed since a given ref. Used by all vault agents.

• obsidian:vault-health

  Analyze vault link health. Detect orphans, deadends, and unresolved links, present a report, and apply fixes upon approval.

• obsidian:vault-ingest

  Analyze provided markdown files against existing vault knowledge. Propose merge, new note creation, or conflict resolution. Apply after user approval.

• obsidian:vault-link

  Find dead-end notes and suggest relevant links based on content similarity. Apply after user approval.

• obsidian:vault-realign

  Detect and fix folder/category frontmatter mismatches. Propose corrections and apply after user approval.

• obsidian:vault-stubs

  Scan for stub/draft notes and generate an enrichment priority report with tracking list.

• obsidian:vault-summary

  Generate a statistics report for a specific vault folder or the entire vault. Shows file counts, tag distribution, task status, and link health.

• obsidian:wordcount

  Count words and characters in any vault note — returns word count, character count, or both.

Subagents (10)

• consistency-checker

  Detect contradictions between vault notes on overlapping topics. Report conflicts with evidence and suggest resolution based on recency and source authority.

• inbox-ingester

  Orchestrate the full Inbox processing pipeline — convert, triage, extract, ingest, index, link, vocabulary — with user checkpoints between steps

• knowledge-enricher

  Actively enrich stub/draft notes by gathering context from vault, source URLs, and web search. Drives the stub → draft → complete lifecycle.

• qmd-refresher

  run qmd update and embed

• vault-classifier

  Identify unclassified notes, suggest tags/properties, and apply upon user approval

• vault-healer

  Analyze vault health — link integrity, note size, and structure — present report, and apply fixes upon user approval

• vault-ingester

  Analyze markdown files against existing vault knowledge, propose merge/create/conflict resolution, apply after user approval. Receives file list from caller.

• vault-linker

  Find dead-end notes and suggest relevant links based on content similarity, apply after user approval

• vault-realigner

  Detect folder/category frontmatter mismatches, propose corrections, apply after user approval

• vault-stub-reporter

  Scan for stub/draft notes, analyze enrichment priority, generate report and tracking list

MCP servers (1)

• qmd

  runs: qmd mcp

Hooks (6)

• SessionStart

  ${CLAUDE_PLUGIN_ROOT}/hooks/ensure-obsidian.sh

• PreToolUse

  ${CLAUDE_PLUGIN_ROOT}/hooks/restrict-vault-path.sh

• PostToolUse

  obsidian reload

• PostToolUse

  ${CLAUDE_PLUGIN_ROOT}/hooks/mark-dirty.sh

• SessionEnd

  ${CLAUDE_PLUGIN_ROOT}/hooks/qmd-sync.sh

• SubagentStart

  ${CLAUDE_PLUGIN_ROOT}/hooks/qmd-sync-f.sh

Similar plugins