# Obsidian > > - **Type:** Skill - **Install:** `agentstack add skill-zealousear-claude-skills-obsidian` - **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior - **Seller:** [ZealousEar](https://agentstack.voostack.com/s/zealousear) - **Installs:** 0 - **Category:** [Productivity](https://agentstack.voostack.com/c/productivity) - **Latest version:** 0.1.0 - **License:** MIT - **Upstream author:** [ZealousEar](https://github.com/ZealousEar) - **Source:** https://github.com/ZealousEar/claude-skills/tree/main/claude-code/obsidian ## Install ```sh agentstack add skill-zealousear-claude-skills-obsidian ``` Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent. ## About # Obsidian Vault Expert The definitive skill for managing an Obsidian vault. Combines direct file operations (always available) with Obsidian CLI v1.12 (indexed graph queries) for a system that is both reliable and intelligent. ## When to Use - Create, read, edit, organize, or delete notes in the vault - Find orphan notes, dead-end notes, or unresolved links - Query backlinks, tags, properties, or tasks across the vault - Run a vault health check or maintenance session - Create notes from templates (Daily Note, Research Note, Project, etc.) - Build or update Maps of Content (MOCs) - Manage plugins (list, install, enable) - Any question about the vault's structure, content, or graph ## When NOT to Use - Writing Obsidian plugins (TypeScript/plugin API) — use normal coding - Content quality review or grammar checking - Media transcription or lecture slide extraction - Knowledge graph theory or visualization ## Invocation ``` /obsidian # Interactive — asks what you need /obsidian health # Full vault health check /obsidian orphans # Find and fix orphan notes /obsidian create research "Title" # Create a research note from template /obsidian search "query" # Search vault content /obsidian backlinks "note path" # Find what links to a note /obsidian tags # List all tags with counts /obsidian daily # Create or append to today's daily note /obsidian moc "Topic" # Create/update a Map of Content /obsidian plugins # List installed plugins ``` ## Architecture: The 3-Tier Hybrid Model ``` User: /obsidian [command] [args] | v [Tier Selection] — Choose the right tool for the job | +---> CRUD (create, read, edit, delete, search content)? | → Tier 1: Direct File Ops (Read/Write/Edit/Glob/Grep) | → Always available, zero dependencies, instant | +---> Discovery (orphans, backlinks, tags, properties, indexed search)? | → Tier 2: Obsidian CLI v1.12 | → Pre-check: pgrep -x Obsidian (must be running) | → Binary: /Applications/Obsidian.app/Contents/MacOS/Obsidian | → IMPORTANT: vault="$VAULT_NAME" MUST be first parameter | → Fallback: Tier 1 (slower but functional) | +---> UI control (open note, trigger search panel)? → Tier 3: obsidian:// URI scheme → open "obsidian://open?vault="$VAULT_NAME"&file=..." ``` ## Vault Configuration Every CLI command below uses `vault="$VAULT_NAME"`, a shell variable you set once. Resolve the vault name and path with this order: 1. If `$VAULT_NAME` is already set in the environment, use it. 2. Otherwise, run `obsidian vault list` and prompt the user to pick one. Export `VAULT_NAME` in the current shell so subsequent commands work. 3. For the absolute vault path, set `$VAULT_PATH` to the directory that contains the `.obsidian/` folder, or walk up from the current working directory until you find one. The example "Vault Details" block below describes one possible vault layout — your PARA folder names, templates, and installed plugins will differ. ## Vault Details (example) - **Path**: `$VAULT_PATH` (set via env var, or detect by locating the `.obsidian/` folder) - **PARA folders** (example layout): 00 Inbox, 01 Projects, 02 Areas, 03 Resources, 09 Systems, 10 School, 99 Archive - **Templates** (example, 12): in `09 Systems/Templates/` — Daily Note, Weekly Review, Project, Research Note, Literature Note, Company Research, Job Application, Learning Plan, Networking Log, Quant Prep, Skill Log, Template Index - **Installed plugins** (example): terminal, calendar, templater-obsidian - **CLI version**: 1.12.1 ## Domain Knowledge Deep reference material lives in the spawner skill. Read these files for detailed patterns, sharp edges, and architectural decisions when needed: ``` ~/.spawner/skills/creative/obsidian-cli/ ├── skill.yaml # Core definition, 13 patterns, 13 domains ├── patterns.md # 11 pattern deep-dives with full examples ├── sharp-edges.yaml # 15 pitfalls with detection patterns ├── sharp-edges.md # 15 sharp edges with Why/Detect/Fix/Prevent ├── collaboration.yaml # Ecosystem, prerequisites, handoffs ├── decisions.md # 11 architectural decisions with rationale ├── anti-patterns.md # 9 anti-patterns with bad/good examples └── validations.yaml # 16 quality gate checks ``` Read the relevant file when you need deep context on a specific topic. ## Command Reference ### /obsidian health — Full Vault Health Check The canonical hybrid workflow. Uses CLI for fast discovery, file ops for fixes. ``` Step 1: Pre-flight checks - Verify Obsidian is running: pgrep -x Obsidian - Verify CLI responds: obsidian vault="$VAULT_NAME" version - If CLI unavailable, warn user and fall back to file-ops-only mode Step 2: Discovery via CLI (Tier 2) - obsidian vault="$VAULT_NAME" orphans total → count orphan notes - obsidian vault="$VAULT_NAME" deadends total → count dead-end notes - obsidian vault="$VAULT_NAME" unresolved total → count unresolved links - obsidian vault="$VAULT_NAME" tags all counts → tag distribution - obsidian vault="$VAULT_NAME" tasks all todo total → open task count Step 3: Structural analysis via file ops (Tier 1) - Glob each PARA folder to count notes per folder - Grep for notes missing frontmatter (no leading ---) - Grep for notes in vault root (should be zero) Step 4: Report - Present findings as a markdown summary - Categorize issues by severity (critical, warning, info) - Offer to fix issues (add links to orphans, create missing notes, etc.) Step 5: Fix (if user approves) - Use Edit to add [[links]] to orphan notes - Use Write to create stub notes for unresolved links - Use Edit to add frontmatter to notes missing it ``` ### /obsidian orphans — Find and Fix Orphan Notes ``` Step 1: CLI discovery obsidian vault="$VAULT_NAME" orphans → List of notes with no incoming links Step 2: For each orphan, analyze with file ops - Read the orphan note - Identify its topic, tags, and content - Search for related notes that should link to it Step 3: Suggest or apply fixes - Add [[links]] in related notes pointing to the orphan - Add the orphan to relevant MOCs - If the orphan is stale, suggest archiving to 99 Archive/ ``` ### /obsidian create [type] "Title" — Create Note from Template ``` Supported types: research → Research Note v1.md → 03 Resources/ project → Project v1.md → 01 Projects/ daily → Daily Note v1.md → 00 Inbox/YYYY-MM-DD.md weekly → Weekly Review v1.md → 00 Inbox/ literature → Literature Note v1.md → 03 Resources/ company → Company Research v1.md → 03 Resources/ job → Job Application v1.md → 01 Projects/ learning → Learning Plan v1.md → 01 Projects/ networking → Networking Log v1.md → 02 Areas/ quant → Quant Prep v1.md → 10 School/ skill → Skill Log v1.md → 02 Areas/ Step 1: Read the template from 09 Systems/Templates/ Step 2: Substitute Templater variables with actual values - → current date - → the provided title Step 3: Check for existing file with Glob (prevent overwrite) Step 4: Write the populated note to the correct PARA folder Step 5: Optionally open in Obsidian via URI ``` ### /obsidian search "query" — Indexed Search ``` Step 1: If CLI available: obsidian vault="$VAULT_NAME" search query="[query]" limit=20 → Fast indexed results in ` variables before writing; wrap in code blocks if documenting 8. **Wiki-links over markdown links** — use `[[Note Name]]` for all internal links ## PARA Folder Selection | Content Type | Folder | |-------------|--------| | Active project with deadline | `01 Projects/` | | Ongoing area of responsibility | `02 Areas/` | | Reference material, topic | `03 Resources/` | | Academic, coursework | `10 School/` | | Quick capture, unsorted | `00 Inbox/` | | Templates, dashboards, config | `09 Systems/` | | Completed, inactive | `99 Archive/` | ## Performance Benchmarks | Operation | CLI v1.12 | File Ops | CLI Speedup | |-----------|-----------|----------|-------------| | Orphan detection | 0.26s | 15.6s | 60x | | Search | 0.32s | 1.6s | 5x | | Backlinks | ~0.1s | Impossible* | -- | | Tag aggregation | ~0.2s | ~3s | 15x | *Backlink queries via file ops require scanning every file for `[[target]]`, missing aliases. ## Source & license This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code. - **Author:** [ZealousEar](https://github.com/ZealousEar) - **Source:** [ZealousEar/claude-skills](https://github.com/ZealousEar/claude-skills) - **License:** MIT Install and usage instructions live in the source repository linked above. ## Pricing - **Free** — Free ## Security capabilities Automated source analysis of v0.1.0 — what this tool can access: - **Network access:** no - **Filesystem access:** no - **Shell / process execution:** no - **Environment & secrets:** no - **Dynamic code execution:** no *"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.* ## Versions - **0.1.0** — security scan: passed — Imported from the upstream source. ## Links - Listing page: https://agentstack.voostack.com/l/skill-zealousear-claude-skills-obsidian - Seller: https://agentstack.voostack.com/s/zealousear - Browse the marketplace: https://agentstack.voostack.com/browse --- Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.