oh-my-second-brain 0.2.0 → 0.3.0

package/adapters/hermes/skills/install/SKILL.md

@@ -1,19 +0,0 @@
----
-name: install
-description: Install Oh My Second Brain Codex/Hermes/Claude host adapters and MCP registration.
----
-
-# oms-install
-
-Use for host lifecycle installation.
-
-```bash
-oms install --runtime <auto|all|claude|codex|hermes> --vault <vault> --yes
-```
-
-For Codex, this installs:
-
-- `~/.codex/rules/oms.md`
-- `~/.codex/skills/oms-*`
-- `~/.codex/plugins/oms`
-- managed `[mcp_servers.oms]` in `~/.codex/config.toml`

package/adapters/hermes/skills/link/SKILL.md

@@ -1,18 +0,0 @@
----
-name: link
-description: Two-path note-linking skill — Path A (pre-write): compose a new note body that already carries [[wikilinks]] to existing term notes; Path B (retrofit): add missing links to existing notes via oms linkify or oms_link_apply.
----
-
-# oms-link
-
-Thin pointer to `core/skills/link`. Requires `OMS_VAULT`.
-
-Link targets are ONLY notes bound to the `term` concept in `vault/.oms/taxonomy.yaml`; their surface forms are the basename and the frontmatter `aliases`. Three engine rules hold in both paths: a link exists only where the term literally appears (surface-anchored — retrieval never invents one); first occurrence per target note only; ambiguity (`ambiguous: true` + `rivalPaths`) is reported, never auto-resolved — ask the user or skip the span. Frontmatter, code, existing links, headings, URLs, and tags are masked out, so a "missing" link inside any of them is correct.
-
-**Path A (pre-write) — use while drafting the note:** Call MCP `oms_link_suggest` on an existing note in the same folder (or run `oms linkify --folder <folder>` in report mode) and read the `targetPath` list — that is the vault's current term vocabulary. Draft the body with those exact surface forms, writing `[[wikilinks]]` inline at the first mention of each term, then commit via MCP `write` (host file edits never touch vault `.md`). Re-check with `oms_link_suggest`: zero candidates means the draft was already fully linked.
-
-NOTE: `0 term note(s) available as link targets` means the vault has no term layer yet — build one with the `wiki` skill first.
-
-**Path B (retrofit) — use when the note already exists:** Run `oms linkify --folder <folder>` first; it reports only and writes nothing. Show the report and get agreement before mutating. For a whole folder use `oms linkify --folder <folder> --apply --yes` — `--apply` without `--yes` refuses before reading any note and writes nothing. For per-note acceptance call `oms_link_suggest { notePath, folder? }` for candidates plus `baseContentHash`, drop unresolved ambiguous ones, then `oms_link_apply { notePath, baseContentHash, candidateIds, folder? }` with the SAME `baseContentHash` and the SAME `folder` scope. Run `oms doctor` afterwards (non-blocking, exits 0).
-
-NOTE: `oms_link_apply` refuses without writing on `note-changed`, `candidate-drift`, or overlapping candidates. Re-run `oms_link_suggest` for fresh offsets instead of retrying stale ids. Every write — CLI or MCP — goes through the capture kernel, so path safety and the concept contract hold. The two paths are mutually exclusive.

package/adapters/hermes/skills/retrieve/SKILL.md

@@ -1,32 +0,0 @@
----
-name: retrieve
-description: Retrieve vault knowledge axis-first using folders, frontmatter, wikilinks, and declared lenses.
----
-
-# oms-retrieve
-
-Use `oms_retrieve_context` first for natural-language retrieval. It combines live
-frontmatter/folder/wikilink graph context with optional OMS native semantic
-lexical/SQLite-vector/HYDE candidates and does not require a warm graph cache. Pass declared
-concept/folder/property/value/wikilink axes when the user's intent gives them.
-
-Use `semanticScope: "global"` for broad vault-wide semantic search. Use
-`semanticScope: "graph"` (**gph mode**) when semantic candidates must stay inside
-the selected OMS graph candidates.
-
-Pass typed sub-queries via `semanticSearches` to combine search strategies in one call:
-`type: "lex"` for BM25 keyword search, `type: "vec"` for SQLite-vector semantic
-search, and `type: "hyde"` for hypothetical-document expansion. Pass `semanticMode`,
-`semanticIntent`, `semanticLex`, `semanticVec`, `semanticHyde`, and
-`semanticMinScore` for additional typed or thresholded control. Pass
-`embeddingSyncBeforeSearch: true` when the native semantic index must be refreshed
-before retrieval, or call `oms_sync_embeddings` explicitly.
-Use read-only `oms_get_document` or `oms_multi_get_documents` to rehydrate selected
-docids, paths, globs, or line ranges. Use `oms_semantic_query`,
-`oms_semantic_status`, `oms_semantic_collections`, or `oms_semantic_contexts` for
-qmd-compatible follow-ups on the native OMS index.
-
-Use `oms_retrieve_by_axis` only for legacy axis-first cache retrieval. Use
-`oms_lazy_load_note` only after selecting candidate notes.
-
-Return lens-shaped fields where possible; do not dump full note bodies unless the user asks or the retrieval task needs body evidence.

package/adapters/hermes/skills/setup/SKILL.md

@@ -1,37 +0,0 @@
----
-name: setup
-description: Adopt an Obsidian markdown vault into the Oh My Second Brain convention and optionally install host MCP integration.
----
-
-# oms-setup
-
-Use when the user wants to initialize Oh My Second Brain for a vault.
-
-Run:
-
-```bash
-oms setup --vault <vault>
-```
-
-For a non-interactive first pass:
-
-```bash
-oms setup --vault <vault> --yes
-```
-
-To interview each folder axis and merge observed note frontmatter into concept schemas:
-
-```bash
-oms setup --vault <vault> --suggest-fields
-```
-
-Then, when host registration is desired:
-
-```bash
-oms install --runtime hermes --vault <vault> --yes
-```
-
-Setup interviews folder intent, concept binding, optional observed fields, and retrieval lenses.
-It rejects lenses that reference unknown fields. Do not modify vault notes during setup.
-Oh My Second Brain writes only `vault/.oms/taxonomy.yaml` and `vault/.oms/concepts/`,
-preserving existing concept files while adding missing defaults or selected observed fields.

package/adapters/hermes/skills/uninstall/SKILL.md

@@ -1,20 +0,0 @@
----
-name: uninstall
-description: Remove Oh My Second Brain host adapter files and MCP registration without deleting vault notes or vault ontology.
----
-
-# oms-uninstall
-
-Preview first:
-
-```bash
-oms uninstall --runtime all --dry-run
-```
-
-Remove host registrations:
-
-```bash
-oms uninstall --runtime all --yes
-```
-
-Never delete vault notes or `vault/.oms/` as part of host uninstall.

package/adapters/hermes/skills/update/SKILL.md

@@ -1,22 +0,0 @@
----
-name: update
-description: Update Oh My Second Brain and refresh installed host adapter registrations.
----
-
-# oms-update
-
-Use when the user wants to update Oh My Second Brain after it has already been installed.
-
-Preview first:
-
-```bash
-oms update --dry-run --runtime all --vault <vault>
-```
-
-Perform the update:
-
-```bash
-oms update --yes --runtime all --vault <vault>
-```
-
-`oms update` checks npm for the latest `oh-my-second-brain` release, installs `oh-my-second-brain@latest` only when `--yes` is present, then reconciles host adapter/MCP registrations through the same install path. Without `--yes`, it prints the planned commands and does not mutate package or host config.

package/adapters/hermes/skills/wiki/SKILL.md

@@ -1,14 +0,0 @@
----
-name: wiki
-description: Two-path wiki skill — Path A: promote compiled concepts from processed/ into the wiki/ query surface; Path B (human authoring): build interlinked terminology notes in vault taxonomy folders via write.
----
-
-# oms-wiki
-
-Thin pointer to `core/skills/wiki`. Requires `OMS_VAULT`.
-
-**Path A — use when `processed/<concept>.md` exists:** Verify the file exists, load `.llmwiki/staleness.json`, promote `processed/→wiki/`, update the ledger, flip cascade backlinks, detect stubs and orphans. Regenerate `wiki/index.md` and append to `wiki/log.md`. Lint — auto-fix index consistency, broken `[[wikilinks]]`, and See-Also sections; report conflicts, orphans, and outdated refs. Only `wiki/` crosses the Obsidian sync boundary; `processed/` is internal.
-
-NOTE: Delete `.llmwiki/staleness.json` to force a full rebuild — every page resets to DIRTY.
-
-**Path B (human authoring) — use when you have only a topic + terms, no compile output:** Ask for topic and term list. Resolve the target folder from `vault/.oms/taxonomy.yaml`. Apply the admission gate: coined/proper terms (only ever mean one specific technical thing) → standalone note; general nouns (meaning shifts by domain) → `## Term` section inside the hub note only. K8s example: `Pod`, `Service`, `Deployment`, `Ingress` → hub sections; `ReplicaSet`, `StatefulSet`, `DaemonSet`, `CronJob` → standalone notes. Deduplicate via MCP `write` (`mode: "create"`) — if the note already exists, reuse the `[[wikilink]]`. Commit coined-term notes first (definition + `## See Also` back-link to hub), then commit the hub/MOC note (general-noun sections + `## See Also` links). Run `oms doctor` (non-blocking, exits 0). Writes to vault taxonomy folders — never to `wiki/`.

package/adapters/hermes/skills/write/SKILL.md

@@ -1,10 +0,0 @@
----
-name: write
-description: Write a vault note through the Oh My Second Brain kernel contract.
----
-
-# write
-
-Call MCP `write`. Do not use host Write/Edit for vault `.md` files.
-
-`mode` is `create`, `append`, or `update`. The kernel owns `.oms` and returns `ask`, `inbox`, `written`, or `rejected`.

package/core/agents/librarian.md

@@ -1,44 +0,0 @@
----
-name: librarian
-description: Places incoming knowledge correctly in the vault, filling frontmatter per the declared concept.
----
-
-# Agent Persona: Librarian
-
-The librarian places knowledge into the vault with precision.
-Every note lands in the right folder, carries the right frontmatter, and respects the declared convention.
-
-## Governing principle
-
-> "A note without declared intent is noise. A note with correct frontmatter is a first-class citizen."
-
-The librarian never improvises a folder or invents a field.
-It reads the convention first, then acts.
-
-## Responsibilities
-
-1. **Concept identification** — match incoming knowledge to a declared concept (`literature`, `inbox`, etc.) by comparing content and purpose against each concept's `intent`.
-2. **Folder resolution** — look up the correct target folder from `vault/.oms/taxonomy.yaml` based on the matched concept.
-3. **Frontmatter construction** — fill every `required: true` field; fill known optional fields; leave undeclared fields untouched.
-4. **Note creation** — write the file at the resolved path; use the concept's naming convention (default: `YYYY-MM-DD-<slug>.md`).
-5. **Write through the kernel** — call MCP `write`. Do not use host Write/Edit for vault `.md` files.
-6. **Post-write validation** — run `npx oh-my-second-brain doctor` (non-blocking) to confirm the new note is clean.
-
-## Decision rules
-
-| Situation | Action |
-|-----------|--------|
-| Concept is ambiguous | Default to `inbox` and note ambiguity in a `status: needs-review` field |
-| Required field value is unknown | Use a sentinel `"UNKNOWN"` string and flag it; never omit the key |
-| Folder does not exist in taxonomy | Capture to `inbox/`; surface the missing taxonomy entry to the user |
-| User provides extra context fields | Preserve them (`additionalProperties: preserve`) |
-
-## Interaction style
-
-- Ask only for information that is truly missing for required fields.
-- Propose the target folder and concept before writing — let the user confirm once.
-- After writing, show the frontmatter block so the user can see what was committed.
-
-## Skill
-
-Invoke via the `write` skill: `core/skills/write/SKILL.md`.

package/core/agents/retriever.md

@@ -1,62 +0,0 @@
----
-name: retriever
-description: Surfaces the right knowledge fields from the vault for a declared retrieval purpose using lenses.
----
-
-# Agent Persona: Retriever
-
-The retriever surfaces knowledge through the vault's declared lenses (인출).
-It never dumps raw notes — it filters to the fields that matter for the stated purpose.
-
-## Governing principle
-
-> "Retrieval is not search. It is projection: show exactly the fields this purpose requires, nothing more."
-
-The retriever reads lens declarations first, then scans notes,
-then returns only the lens-selected fields.
-
-## Responsibilities
-
-1. **Purpose identification** — clarify the user's retrieval goal (synthesize, audit, plan, review, etc.).
-2. **Lens matching** — find the declared lens on the relevant concept whose `intent` best matches the purpose.
-3. **Field projection** — for each matching note, return only the fields listed in the lens; discard the rest.
-4. **Grouping** — present results grouped by concept/folder for orientation.
-5. **Citation** — always include `source-url` (or equivalent) when the concept carries it, even if the lens does not explicitly list it.
-
-## Lens resolution algorithm
-
-```
-1. Parse user purpose → keyword(s)
-2. For each concept in vault/.oms/concepts/*.yaml:
-     for each lens in concept.lenses:
-       score = semantic_overlap(purpose, lens.intent)
-3. Select lens with highest score
-4. Collect all notes in concept's folder
-5. Project each note's frontmatter to lens.fields
-6. Return projected results
-```
-
-In v0, step 2 is agent judgment (no runtime engine). The retriever reads the YAML manually.
-
-## Output format
-
-```markdown
-## [Concept: literature] — Lens: synthesis
-
-| title | thesis | source-url |
-|-------|--------|------------|
-| Attention Is All You Need | Self-attention replaces recurrence | https://arxiv.org/abs/1706.03762 |
-```
-
-## Decision rules
-
-| Situation | Action |
-|-----------|--------|
-| No matching lens | Fall back to all `required: true` fields |
-| Note missing a lens field | Show `—` for that cell; do not skip the note |
-| Multiple concepts match purpose | Return results from all matching concepts, grouped |
-| No notes in concept folder | Report "0 notes found under `<folder>/`" |
-
-## Skill
-
-Invoke via the `retrieve` skill: `core/skills/retrieve/SKILL.md`.

package/core/skills/compile/CHANGELOG.md

@@ -1,24 +0,0 @@
-# compile — CHANGELOG
-
-## 0.1.0 — 2026-06-13
-
-Initial release (M2 compile engine).
-
-### Attribution
-
-- **SHA-incremental pattern (Steps 2-4)**: bstack `terminology` skill — self-authored, no external restriction.
-- **2-step Chain-of-Thought (analysis → synthesis)**: nashsu/llm_wiki — GPL-3.0, IDEA-ONLY. No verbatim code. Algorithm description absorbed and reimplemented from scratch. See `src/engine/compile/ACKNOWLEDGMENTS.partial.md`.
-- **2-phase separation (extract-without-writing → generate)**: atomicstrata — license TBD, CONCEPT-ONLY. No verbatim code.
-- **Cascade return (write → affected_backlinks)**: lucasastorian/llmwiki — terminology absorbed, concept-only.
-- **Provenance grading**: self-authored (bstack second-brain design).
-
-### Files
-
-- `src/engine/compile/types.ts` — local compile types
-- `src/engine/compile/sha-cache.ts` — SHA-256 fingerprint + disk cache
-- `src/engine/compile/provenance.ts` — grade resolution + synthesis weighting
-- `src/engine/compile/cot.ts` — nashsu 2-step CoT (GPL idea-only reimplementation)
-- `src/engine/compile/phases.ts` — atomicstrata 2-phase separation (concept-only)
-- `src/engine/compile/cascade.ts` — lucasastorian cascade backlink return
-- `src/engine/compile/worker.ts` — stateless compile worker (entry point)
-- `src/engine/compile/*.test.ts` — co-located vitest suites

package/core/skills/compile/SKILL.md

@@ -1,57 +0,0 @@
----
-name: compile
-version: 0.1.0
-description: Agent-guided concept compile — synthesize a concept wiki page from source materials with incremental skip, two-phase write separation, cascade backlinks, and provenance-weighted context.
-trigger: /compile
-tags: [compile, wiki, synthesis, sha, incremental, second-brain, oms]
----
-
-## compile
-
-Synthesize a concept page from source materials. There is no `oms compile` CLI
-and no TypeScript compile worker — follow this recipe as the agent.
-
-### When to use
-
-Use `compile` when you need to synthesize a concept wiki page from one or more
-source materials. Skip rewrite when the material fingerprint is unchanged.
-
-### Inputs
-
-- `concept` — human-readable concept name (used in prompts)
-- `materials` — list of `{ path, text, grade }` items (grade from folder→grade map)
-- backlinks — wiki pages that already link to this concept (empty if unknown)
-- `conceptId` — stable cache key (vault-relative path recommended)
-
-Gather materials with MCP `oms_retrieve_context` or `oms_get_document`.
-
-### Output
-
-- `body` — synthesized Markdown with `[[wikilinks]]` (empty when skipped)
-- `sha` — SHA-256 fingerprint of the input materials
-- `provenance` — grades of contributing materials
-- `affected_backlinks` — vault paths of wiki pages that link to this concept
-
-### Recipe
-
-1. **Phase A** — load and grade all source materials (pure read, no vault mutation).
-2. Fingerprint the materials. If `{dotLlmwiki}/sha-cache.json` already has the
-   same SHA for `conceptId`, skip the rest.
-3. **Phase B** — write the synthesized body to the `processed/` tier only.
-4. Pass `affected_backlinks` to the wiki skill so those pages can be marked stale.
-
-### Phase constraints
-
-- Phase A and Phase B never overlap in one execution context.
-- Phase B writes to `processed/` ONLY — never to `wiki/` directly.
-- Promotion from `processed/` to `wiki/` is the wiki skill's responsibility.
-
-### Provenance weighting
-
-Materials are sorted authored > curated > external-raw in synthesis context.
-Authored materials are labelled `[AUTHORED — preserve individual voice]` in the prompt.
-
-### SHA cache
-
-Cache location: `{dotLlmwiki}/sha-cache.json` (never synced, never committed).
-Delete `sha-cache.json` to force full recompile of all concepts.

package/core/skills/define/SKILL.md

@@ -1,59 +0,0 @@
----
-name: define
-description: Grow the vault convention by adding or editing a metadata field on a concept.
----
-
-# Skill: define
-
-Extend your Oh My Second Brain convention field-by-field.
-Each frontmatter key is a **unit of convention**: it carries a declared `intent`,
-type, and optional rules (`required`, `normalize`, `immutable`).
-
-## Entry point
-
-```bash
-oms define
-```
-
-This is the intended user-facing command (roadmap: interactive runtime not built yet).
-In v0, `setup` and `doctor` are the REAL CLI commands today.
-Use `define` as **agent-guided convention editing** until the interactive runtime ships.
-
-## What this skill does (agent-guided, v0)
-
-1. Ask the user: which **concept** are they extending? (e.g. `literature`, `inbox`)
-2. Ask: what is the new **field name**? (snake-case, e.g. `thesis`)
-3. Ask: what is this field's **intent** — why does this knowledge live here?
-4. Ask: **type** (`string` | `string[]` | `date` | `url` | `boolean`), **required** (yes/no).
-5. Optionally ask: `normalize` (e.g. `lowercase`), `immutable` (lock after creation).
-6. Open `vault/.oms/concepts/<concept>.yaml` and append the new field entry.
-7. Run `oms doctor` to validate existing notes against the updated schema (exits 0).
-
-## Convention YAML shape (one field entry)
-
-```yaml
-fields:
-  - name: thesis
-    type: string
-    required: false
-    intent: "The central claim this source makes"
-```
-
-## Growing lenses
-
-After adding a field, consider whether it belongs in an existing lens (`synthesis`, `audit`)
-or warrants a new lens. Edit the `lenses:` block in the same concept YAML.
-
-## Where the convention lives
-
-All convention data is **user-owned** at `vault/.oms/`:
-```
-vault/.oms/
-  concepts/
-    literature.yaml
-    inbox.yaml
-  taxonomy.yaml
-```
-
-Oh My Second Brain ships defaults (from `core/ontology/`); `oms setup` copies them into the vault.
-The user then edits them at will — Oh My Second Brain only enforces, never overwrites.

package/core/skills/distill/CHANGELOG.md

@@ -1,39 +0,0 @@
-# Distill Skill — Changelog & Attribution
-
-## [0.1.0] — 2026-06-13
-
-### Added
-- `src/engine/distill/types.ts` — local distill types (CleanRoomSpec, DistillTarget,
-  AnalyzerResult, DistillPattern, DistillRisk, DistillAttribution). Isolated from
-  shared `src/engine/types.ts` per parallel-build safety rule.
-- `src/engine/distill/clean-room.ts` — pure function `prepareCleanRoom()` returning
-  a CleanRoomSpec; `snapshotSha()` and `detectMutation()` for mutation-detection.
-- `src/engine/distill/analyzer.ts` — `AnalyzerProvider` seam (mirrors M1
-  EmbeddingProvider injection pattern); `createStubAnalyzerProvider()` deterministic
-  stub; `runAnalysis()` orchestration with schema validation.
-- `src/engine/distill/report.ts` — `generateReport()` producing exactly 3 sections:
-  §1 Patterns (ranked by absorb_confidence), §2 Risks, §3 Attribution.
-- Co-located tests: `clean-room.test.ts`, `analyzer.test.ts`, `report.test.ts`,
-  `smoke.test.ts` — 85 tests, all green.
-- `core/skills/distill/SKILL.md` — /skillify structure, 5-key frontmatter.
-- `src/engine/distill/ACKNOWLEDGMENTS.partial.md` — partial attribution fragment.
-
-### Absorbed patterns (recorded for ACKNOWLEDGMENTS.md)
-
-| Pattern | Source | License | Notes |
-|---------|--------|---------|-------|
-| Provider injection seam (AnalyzerProvider mirrors EmbeddingProvider) | M1 embed module — self-authored (oh-my-secondbrain) | — | Internal ref; no external restriction |
-| Clean-room as pure data structure (spec-only, caller spawns) | Plan design — self-authored | — | R6 isolation principle |
-| SHA-256 mutation detector | Node.js crypto — standard library | Node.js | Standard API; no absorption needed |
-| Known-pattern signatures in stub | docs/research/embedding-pipeline-patterns-mining.md | — | Self-authored absorption ledger |
-| Known-pattern signatures in stub | docs/research/graphify-graph-implementation-mining.md | — | Self-authored absorption ledger |
-
-### Design rationale
-- **Standalone** (R16): distill is not a router leaf. It does not import from
-  `src/engine/compile/` or `src/engine/wiki/`. It works on any target.
-- **Vault-agnostic**: OMS_VAULT env not required. No hardcoded paths.
-- **Stateless** (R2): no daemon, no watcher, no persistent process.
-- **Provider injection**: mirrors M1 EmbeddingProvider pattern so real LLM
-  clients can be injected in production without changing the orchestration layer.
-- **Mutation detector**: verifies R2/R6 — main system state must be unchanged
-  before and after every distill run.

package/core/skills/distill/SKILL.md

@@ -1,62 +0,0 @@
----
-name: distill
-version: 0.1.0
-description: >
-  Standalone meta-absorption skill. Analyzes any target (repo, document, skill,
-  concept) in a clean-room context and produces a structured absorption report
-  with patterns, risks, and attribution. Vault-agnostic — OMS_VAULT not required.
-trigger: /distill
-tags:
-  - absorption
-  - red-team
-  - clean-room
-  - analysis
-  - standalone
----
-
-# Distill
-
-Run adversarial absorption analysis on a target and produce a structured report.
-There is no TypeScript distill engine — follow this recipe as the agent.
-
-## Usage
-
-```
-/distill <target-path-or-text>
-```
-
-## Steps
-
-1. Load the target as inert read-only text. The target is NEVER executed — it
-   is treated as data only.
-2. Snapshot a content hash of the working tree (or target file) before analysis.
-3. Run red-team adversarial analysis. Use a real model for production; keep
-   the pass read-only.
-4. Generate the absorption report. The report has exactly 3 sections:
-   §1 Patterns, §2 Risks, §3 Attribution.
-5. Verify the hash after analysis equals the hash before. Stop if anything
-   was mutated — distill must be fully stateless.
-6. Return the report string. Write it only if the caller explicitly requests it.
-   No vault write, no code mutation happens inside distill.
-
-## Target types
-
-- **Repo mining doc** — load `docs/research/*-mining.md` as inert text.
-- **Flat document** — any markdown, text, or code file loaded as a string.
-- **Live repo** — read files from disk as text; do not execute them.
-
-## Output
-
-The report is the only output. It contains:
-
-- **§1 Patterns** — ranked by absorb_confidence (highest first), with file:line citations.
-- **§2 Risks** — ranked by severity (critical → high → medium → low).
-- **§3 Attribution** — repo, URL, and license note for `ACKNOWLEDGMENTS.md`.
-
-## Constraints
-
-- Standalone: does not depend on compile or wiki.
-- Vault-agnostic: `OMS_VAULT` environment variable is NOT required.
-- Stateless: no daemon, no watcher, no `setInterval`.
-- Clean-room: target content is inert text — never executed.
-- No vault write occurs inside distill — the caller decides what to do with the report.

package/core/skills/doctor/SKILL.md

@@ -1,49 +0,0 @@
----
-name: doctor
-description: Validate vault notes against the Oh My Second Brain convention and report violations (REAL today).
----
-
-# Skill: doctor
-
-Check every note in your vault against its declared concept schema.
-This skill is **REAL in v0** — the CLI command is fully implemented.
-
-## Shell-out
-
-```bash
-oms doctor [--vault <path>]
-```
-
-- `--vault <path>` — path to your Obsidian vault root (default: current directory).
-- Always **exits 0** in v0 (non-blocking; `onViolation: warn`).
-
-## What the command does
-
-1. Loads `vault/.oms/` and `vault/.oms/taxonomy.yaml`.
-2. Walks every `.md` note in the vault.
-3. For each note, resolves its **concept** from the taxonomy folder binding.
-4. Calls `validateFrontmatter(frontmatter, concept)` and collects `Violation[]`.
-5. Prints a per-note summary of violations (field, rule, message).
-6. Prints total counts: notes checked, violations found.
-7. Exits 0 regardless of violation count (v0 is advisory only).
-
-## Violation rules
-
-| Rule | Description |
-|------|-------------|
-| `required` | A field declared `required: true` is missing from the note's frontmatter |
-| `type` | The field value does not match the declared type |
-| `immutable` | Field changed after initial set — **suppressed in v0** (no baseline available) |
-
-Undeclared frontmatter fields are **never** reported as violations
-(`additionalProperties: preserve`).
-
-## Recommended usage
-
-Run `oms doctor` after:
-- Any `oms setup` run
-- Adding a new field via `oms define`
-- Bulk-editing notes
-
-Integrate into CI by adding `npx oms doctor --vault ./vault` to your workflow.
-It will never fail the build in v0.