agentleFS
synthetic-reconsynthetic-recon/atlas-obsidian-vault/.agents/skills/vault-init/SKILL.md

vault-init

Skill2 starsChanged 7 months agoMITView on GitHub

Initialize a new research vault with the standard directory structure, Obsidian configuration, AGENTS.md, templates, custom callouts, hot cache, delta tracking manifest, and utility scripts. Use when the user asks to create a new vault, initialize a vault, or set up a new research knowledge base.

3 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read AGENTS.md (or CLAUDE.md) to understand conventions

    line 81

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Vault Init Skill

Initialize a new research vault with the standard directory structure, Obsidian configuration, and agent instructions following the LLM-wiki pattern.

Usage

".agents/skills/vault-init/vault-init.sh" "<path-to-new-vault>"

Example:

".agents/skills/vault-init/vault-init.sh" "/path/to/my-new-research-vault"

Vault Structure Created

vault-name/
├── .obsidian/                          # Obsidian configuration
│   ├── app.json
│   ├── appearance.json                 # Enables wiki-callouts CSS snippet
│   ├── core-plugins.json
│   ├── graph.json
│   ├── workspace.json
│   └── snippets/
│       └── wiki-callouts.css           # Custom callouts: contradiction, gap, key-insight, stale
├── _templates/                         # Page templates per type
│   ├── Source.md
│   ├── Concept.md                      # Includes Counter-arguments & Data Gaps
│   ├── Entity.md
│   ├── Question.md
│   └── Output.md
├── bin/                                # Utility scripts (executable)
│   ├── vault-health.sh                 # Severity-tiered lint
│   ├── cross-linker.sh                 # Finds unlinked mentions
│   └── yt-ingest.sh                    # yt-dlp wrapper for video ingest
├── methodology/                        # Canonical reference documents
│   ├── llm-wiki.md
│   └── 2026-04-02T204221.000Z - Thread by @karpathy.md
├── raw/                                # Immutable source documents
│   ├── assets/                         # Local images/attachments
│   └── .manifest.json                  # Delta tracking (source hashes → wiki pages)
├── AGENTS.md                           # Agent schema (canonical)
├── CLAUDE.md                           # Symlink → AGENTS.md
├── Vault Health.base                   # Obsidian Bases dashboard
├── hot.md                              # Session continuity cache
├── index.md                            # Navigation hub
└── log.md                              # Append-only operation record

Note: Wiki content folders (Sources/, Concepts/, Entities/, Questions/, Outputs/) are created on demand during first ingest — not pre-created. This matches the AGENTS.md principle of minimal scaffolding.

What's Included

  • Agent schema (AGENTS.md): Full instructions for ingest, query, lint workflows plus conventions for cross-linking, provenance, source conflicts, page lifecycle, frontmatter schema, and more
  • Custom callouts CSS: Styled callouts for contradictions, gaps, key insights, and stale claims
  • Page templates: Starter scaffolds with correct frontmatter for each page type
  • Hot cache (hot.md): Session continuity file for cross-session context
  • Delta tracking (raw/.manifest.json): Hash-based source tracking to prevent re-ingesting
  • Vault Health dashboard (Vault Health.base): Obsidian Bases view with multiple lenses
  • Utility scripts: vault-health.sh (severity-tiered lint), cross-linker.sh (unlinked mentions), yt-ingest.sh (video transcripts)
  • Methodology docs: The LLM-wiki pattern reference and Karpathy thread
  • Obsidian config: Bases, Canvas, Properties, Backlinks, Graph pre-enabled

What's NOT Copied

  • skills-lock.json — each vault manages its own skills
  • .agents/ contents — skills vendored per vault
  • .claude/settings.json — each project manages its own Claude Code config
  • Any raw source files or existing wiki pages — start fresh

After Initialization

  1. Open the vault in Obsidian
  2. Read AGENTS.md (or CLAUDE.md) to understand conventions
  3. Drop the first source into raw/ and begin ingesting (follow the Ingest workflow in AGENTS.md)
  4. The hot cache, index, and log will populate as you work

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.