# Skills Catalog

This repo now includes the legacy `videoclaw/skills` set, imported into the
clean-room tree so the workflow surface is available locally instead of only in
the old repo.

## Start Here

The imported library is no longer treated as a flat bag of equally-preferred
entry points.

Use these canonical surfaces first:

| Goal | Start With | Then Specialize Into |
|---|---|---|
| "Help me make a video" — guided, novice-facing, menu-driven | `concierge` (speaks as **Clawbot**, the VideoClaw mascot; `clawbot` is a persona alias) | routes into any lane: `video-framework`, `brand-presenter`, `brand-agency`, `character-ad`, `ugc`, `movie-director`, `multi-shot-prompt`, `video-clone-ad` |
| Generic video request | `video-framework` | `video-storyboard`, `video-analyze-template`, `video-clone-ad`, `video-replicator`, `movie-director`, `video-production-handoff`, `video-review-ui-qa`, `video-portfolio-ops`, `video-release-readiness`, `higgsfield-generate`, `ugc` |
| Narrated presenter / host-led deck video | `brand-presenter` | `davendra-presenter`, `nex-presenter`, `bunty` |
| Kids sung nursery rhyme / character series episode | `rhyme-factory` (songs) or `garden-days` (stories) | both read a character pack from `$VCLAW_WORKSPACE/packs/<name>.pack` — one pack per character/channel, never a skill per character |
| Narrated brand explainer film (house motion-graphics style) | `brand-explainer` | one shared deterministic engine + a `.brand` pack from `$VCLAW_WORKSPACE/packs/` — one pack per film, never a copied engine; style doctrine lives in `mograph` |
| Rap / avatar music video from a portrait photo | `rap-avatar-mv` | reads a `.rap.pack` from `$VCLAW_WORKSPACE/packs/`; `dance-breakdown` supplies the per-shot choreography breakdown its shot-refs stage consumes |

### `/slash` front doors

A skill that should be invokable by a `/name` slash command (or trigger phrase)
needs a **front-door pointer** — a thin `SKILL.md` in your local Claude Code
skills-pointer directory (the gitignored, per-machine `.claude` config), which
is where the Skill tool looks. It is NOT created by adding the canonical
`skills/<name>/SKILL.md` alone. `/garden-days` was silently invisible for
exactly this reason until its pointer was added by hand (2026-07-22).

Do not hand-create pointers. Declare the front door once in
`scripts/skill-frontdoors.json` (name, aliases, triggers, one-line summary),
then run **`npm run frontdoors:install`** — it materializes the thin pointer(s)
and a trigger block in your local config, idempotently. `npm run
frontdoors:check` reports drift. Editing a front door's summary or triggers
later is enough: the installer refreshes a stale pointer and any trigger
block it wrote (they carry a `_(generated by …)_` stamp). A block written by
hand is never rewritten — it is reported as a note; delete it to regenerate.
A pointer the Skill tool has re-materialized as a full copy is left alone. A repo test (`skill-frontdoors.test.ts`, in
`npm test`) keeps the declaration honest — every declared front door must point
at a real `skills/<name>/SKILL.md`. (This is a selected slash surface, not the full catalogue. Local skills such as
graphify, improvement-run and higgsfield-generate are currently not declared here.)

Compatibility/deep-reference surfaces remain available, but the preferred UX is:

1. start at a canonical entry point
2. specialize only when the mode is clearly known
3. keep personal or legacy names as aliases rather than first-choice workflows

## Video Production Skills

- `garden-days` — Baby Ram & Sita three-lane series factory (writers → GB identity images → Runway/Higgsfield/Veo queues → automated finish)
- `rhyme-factory` — sung nursery-rhyme YouTube videos on top of `garden-days` (kie.ai Suno song → sing-along captions → title-card/end-card polish → thumbnail + upload pack); character config via `$VCLAW_WORKSPACE/packs/`
- `dance-breakdown` — a reference dance video → a measured per-shot choreography breakdown (its own scene detection, each shot's footage sent inline to a multimodal model): one moves sentence for a window that sings, timecoded blocks for a window with no voice, framing/camera/subject facts; consumed by `rap-avatar-mv`'s shot-refs stage
- `brand-explainer` — narrated 60–70s brand explainer films (house motion-graphics style: GB illustration sheet → typed command cards → MY TAKE → QR outro → Rachel VO + ducked bed); ONE shared deterministic engine, film config via `$VCLAW_WORKSPACE/packs/<brand>.brand`
- `ad-intel` — competitor ads torn down with our own tools, no paid scraping API (Gemini analysis may be billed): the public Meta Ad Library fetched with the cloak browser (or one public video URL) → whisper → Gemini → one artifact per ad, a brand teardown (angle clusters by the `ugc` belief vocabulary, verbatim hooks, longest-running / most-repeated) and a review page; feeds `brand-agency` Phase 2 and `ugc` research
- `video-framework`
- `video-replicator`
- `video-storyboard`
- `video-analyze-template`
- `video-clone-ad`
- `video-production-handoff`
- `video-review-ui-qa`
- `video-portfolio-ops`
- `video-release-readiness`
- `movie-director`
- `ai-film-director` — platform-aware scene director: minimal idea (+ character / first-frame / last-frame refs) → a ten-section production prompt for Seedance 2.5 or Flow/Veo, with handoff-continuous multi-scene sequences; `/film-director` and `/davs-ai-film-director` are aliases
- `higgsfield-generate`
- `seedance-prompts`
- `mograph`
- `ugc`
- `character-library`
- `character-creator`
- `video-post`
- `youtube-audio`
- `brand-presenter`
- compatibility aliases:
  - `davendra-presenter`
  - `nex-presenter`
  - `bunty`

## Workflow Skills

Generic orchestration skills that duplicate the operator's global plugin set
(`autopilot`, `ralph`/`ralph-init`/`ralplan`, `team`, `cancel`, `trace`, `hud`,
`git-master`, `code-review`, `security-review`, and the removed-tooling
`omx-setup`) have been culled from this repo — use the global versions. The
remaining repo-local workflow skills:

- `concierge`
  - persona alias: `clawbot`
- `improvement-run` — whole-codebase improvement pass (parallel audit → adversarial verification → plan gate → batched PRs with fresh verifiers)
- `deepsearch`
- `graphify` — codebase knowledge-graph navigation (query/path/explain over a built graph; external graphify CLI + graphify-out/)
- `web-clone`
- `ai-slop-cleaner`

## Import Notes

1. The imported skill tree preserves legacy names for compatibility.
2. Personal/brand-specific presenter skills are kept as aliases over the shared
   `brand-presenter` surface so the repo stays generic while preserving
   discoverability.
3. `video-framework` and `brand-presenter` are the canonical entry points for
   broad video and presenter requests; deeper skills now sit underneath them as
   specialist or compatibility surfaces.
4. `character-library` is the hygiene and repair companion to
   `character-creator`; creation stays on the creator skill while audit, patch,
   and delete operations live in the library lane.
5. `video-post` closes the loop after render with verification, platform
   variants, thumbnails, and archive helpers.
6. `video-production-handoff` is the operator handoff lane for Review UI,
   review-autopilot, canonical `review-report.json` truth, and release-ready
   verification before publish.
7. `video-review-ui-qa`, `video-portfolio-ops`, and
   `video-release-readiness` split the production surface into browser QA,
   slate management, and release handoff so advanced commands remain intact but
   operators get clear workflows.
8. `higgsfield-generate` is an external bridge over Higgsfield's MIT-licensed
   public skills and official CLI. Keep it optional and thin; do not vendor the
   upstream setup script or make Higgsfield a required runtime dependency.
9. Local repo paths should be preferred over old `.claude/...` path assumptions
   when adapting or extending these skills further.

## Machine-Readable Catalog

For tooling and maintenance, the repo now also carries:

1. `skills/catalog.json`

It records the imported skill ids, rough category, alias relationships, and the
new canonical-entry/specialization structure so the skill surface can be
audited without scraping markdown.