obsidian-apprentice-rules
Skill9 starsChanged 10 days agoMITView on GitHub
Use whenever importing, organizing, or writing into the owner's Obsidian vault; enforces Kepano-aligned apprentice constraints.
In this repository6 other files it gives its agents.
- digital-garden-onboardingSkill
- flomo 推文入库Skill
- getting-startedSkill
- Obsidian 批量剪藏Skill
- orphan-notesSkill
- related-notesSkill
The file
Files
Obsidian Apprentice rules
Apply on every vault task. Source mindset: Kepano · How I use Obsidian.
Hard rules
- File over app — real
.mdonly; standard Markdown; no invented syntax. - Sparse folders — new notes default to vault root unless the owner says otherwise;
Clippings/for other people's writing; your authored output only underAI outputs/(or their named AI-output folder); no silent mass re-filing. - Wikilinks first — keep and complete
[[wikilinks]]; never peel concepts into bare words without brackets. - Unresolved links OK — leave
[[Name]]as breadcrumbs; do not batch-create empty stub notes for missing targets. - Clipping quality gate — import only usable full text. Before writing: spot-check ≥1 note for long body, presence of
[[ ]], and that concept notes can grow backlinks. Reject "summary + full text on X" shells. - Don't understand for them — you may scan, clean, and file; elevating ideas into claims and random revisits stay with the owner.
- Minimal properties — add YAML only when needed; don't invent experimental fields.
- Ask first — major deletes and batch moves need the owner's go-ahead.
- Lived-experience gate for publishable drafts — when turning notes into outward writing, keep only what the owner actually did; don't invent experience.
Import checklist
- Spot-check ≥1 note: full body, not a redirect shell
- Concepts appear as
[[Concept]], not plain tokens - Concept notes link back with
[[Article title]](or backlinks will appear after save) - Destination folder matches map (others → Clippings; your output → AI outputs)
- No mass empty stubs for missing links
Anti-pattern
Importing open-source summary shells and converting Concepts' [[links]] to plain text — backlink graph goes to zero. Never repeat.
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
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.

