AdoptSAFE
A complete starter kit for an Obsidian + Claude Code personal knowledge management system.
Overview
A complete starter kit for an Obsidian + Claude Code personal knowledge management system.
855f65f39078OBSERVED · 2026-10-08What it tells the agent
The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.
---
name: adopt
description: Scaffold the PKM system onto an existing Obsidian vault. Scans your vault structure, maps folders interactively, and generates configuration — no template required.
allowed-tools: Read, Write, Edit, Glob, Grep, Bash, AskUserQuestion
model: sonnet
user-invocable: true
---
# Adopt Skill
Bring Your Own Vault (BYOV) — install the Claude PKM system onto an existing Obsidian vault.
## Usage
```
/adopt # Run from the root of your existing Obsidian vault
```
## When to Use
- You have an existing Obsidian vault and want to add the Claude PKM system
- You don't want to start from the vault template
- You want to keep your current folder structure
## Phase 1: Scan Vault Structure
Analyze the existing vault to understand its organization.
### Steps
1. **List top-level directories** using `ls`, excluding system dirs (`.obsidian`, `.git`, `.claude`, `.trash`, `.claude-plugin`)
2. **For each directory**, gather signals:
- Count `.md` files (using Glob)
- Check for date-named files (`YYYY-MM-DD*.md`) — indicates daily notes
- Grep for goal/review/template keywords in filenames and content
- Check for existing `CLAUDE.md` in subdirs — indicates projects
3. **Detect organization method** based on signals:
- **PARA**: Folders named Projects, Areas, Resources, Archives
- **Zettelkasten**: Numeric-prefixed notes, heavy wiki-linking, flat structure
- **Johnny Decimal**: `00-09`, `10-19` style folder names
- **LYT (Linking Your Thinking)**: Folders named Atlas, Calendar, Cards, Extras, Sources
- **Flat**: Few folders, most files at root
- **Custom**: None of the above patterns match
4. **Present findings** to the user:
```
Vault scan complete!
Found 342 notes across 8 folders:
Daily/ → 180 notes (date-named — likely daily notes)
Projects/ → 45 notes (has CLAUDE.md files — likely projects)
Goals/ → 12 notes (contains goal keywords)
Templates/ → 8 notes (contains template keywords)
Archive/ → 67 notes
Inbox/ → 15 notes
Resources/ → 10 notes
Meeting Notes/ → 5 notes
Detected method: PARA-like structure
```
## Phase 2: Map Folders to Roles
Use AskUserQuestion to confirm or correct the detected mappings.
### Roles to Map
Each role maps a PKM function to a folder in the user's vault:
| Role | Purpose | Detection Signal |
|------|---------|-----------------|
| Daily Notes | Daily journal entries | Date-named files (`YYYY-MM-DD`) |
| Goals | Goal cascade (3-year → weekly) | Files with goal/review keywords |
| Projects | Active projects | Subdirs with CLAUDE.md or project keywords |
| Templates | Reusable note structures | Files with template keywords or in Templates/ |
| Archives | Completed/inactive content | Folder named Archive(s) or with old dates |
| Inbox | Uncategorized captures | Folder named Inbox, or files tagged #inbox |
### Interactive Mapping
For each role, ask the user to confirm or correct:
**Question format (use AskUserQuestion):**
- "Which folder holds your **daily notes**?"
- Options: detected candidate(s), "I don't have one (create it)", "Skip — I don't use this"
- For optional roles (Inbox), include "Skip" as a default
**Edge cases:**
- **Existing CLAUDE.md at root**: Ask the user — back up as `CLAUDE.md.backup` or merge content
- **No candidate for a role**: Offer to create the folder
- **Multiple candidates**: Present all and let the user choose
### Save Mappings
Store the folder mapping for use in later phases:
```
dailyNotes → "Daily"
goals → "Goals"
projects → "Projects"
templates → "Templates"
archives → "Archive"
inbox → "Inbox" (or null if skipped)
```
## Phase 3: Personalize Preferences
Ask the same 4 questions as `/onboard`:
**Question 1: Your name**
- "What should I call you?"
- Used for personalized prompts and greetings
**Question 2: Preferred review day**
- "What day do you prefer for your weekly review?"
- Options: Sunday (Recommended), Saturday, Monday, Friday
- Used by `/review` auto-detection and session-init nudges
**Question 3: Primary goal areas**
- "Which areas are most important to you right now? (Pick 2-4)"
- Options: Career & Professional, Health & Wellness, Relationships, Personal Growth
- Also offer: Financial, Creativity & Fun, Learning, Other
- multiSelect: true
- Used to customize goal template suggestions
**Question 4: Work style**
- "How do you prefer Claude to interact?"
- Options: Direct and concise (Recommended), Coaching and challenging, Detailed and thorough, Minimal — just do the task
- Sets output style preference
## Phase 4: Generate Configuration
### 4a. Write `settings.json`
Write `.claude/settings.json` with permissions scoped to the user's actual folders:
```json
{
"permissions": {
"allow": [
"Read",
"Write **/{mapped-daily}/**",
"Write **/{mapped-goals}/**",
"Write **/{mapped-projects}/**",
"Write **/{mapped-templates}/**",
"Edit **/{mapped-daily}/**",
"Edit **/{mapped-goals}/**",
"Edit **/{mapped-projects}/**",
"Glob",
"Grep"
]
}
}
```
Replace `{mapped-*}` with actual folder names from Phase 2.
### 4b. Write root `CLAUDE.md`
Generate a root `CLAUDE.md` that describes the user's **actual** vault structure. Use the same format as the template's CLAUDE.md but with:
- Their folder names in the Directory Structure table
- Their goal areas in the System Purpose section
- Their actual skills table (same as template)
- Cascade section adapted to their folder names
If the user had an existing `CLAUDE.md`, merge their content into the appropriate sections (preserve their mission statement, custom conventions, etc.).
### 4c. Write `vault-config.json`
Write `vault-config.json` in the vault root:
```json
{
"name": "User's name",
"reviewDay": "Sunday",
"goalAreas": ["Career & Professional", "Health & Wellness"],
"workStyle": "Direct and concise",
"setupDate": "2026-02-Trust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | NA |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- none-observed
- Network
- none-observed
- Shell
- none-observed
- Dependencies
- pinned
- Secrets in source
- none-found
Findings (0)
No findings outside the package's declared scope.
Gates applied: no_behavioural_pass.
855f65f39078full audit observations/trust-audit/skill/ballred__adopt.json · Report an issue / request a re-scanAudit history
Every audit this skill has had.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-08 | 855f65f39078 | SAFE | B | 89 | first audit |
Questions
What does the Adopt skill do?
A complete starter kit for an Obsidian + Claude Code personal knowledge management system.
Is Adopt safe to install?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean skill reads B.
What can Adopt access on my machine?
The audit observed no filesystem, network or shell use at all in its source.
How current is this page?
The grade is for one exact copy of the source (855f65f39078), read on 2026-10-08. The repository is watched, and a new audit runs when it changes — this is the first audit.