@amg555/frameforge-mcp 2.7.6 → 2.7.7

package/.agent/skills/ai-video-editor/styles/warm-canvas/style.json

@@ -1,14 +0,0 @@
-{
- "id": "warm-canvas",
- "title": "Warm Canvas",
- "family": "Talking Head",
- "fonts": [
-  "560 60px \"Fraunces\"",
-  "600 30px \"Fraunces\"",
-  "500 20px \"Inter\"",
-  "600 20px \"Inter\""
- ],
- "demo_duration": 32.6,
- "demo_still": 18.9,
- "has_aroll": true
-}

package/.agent/skills/ai-video-editor/styles/warm-canvas/style.md

@@ -1,37 +0,0 @@
-# Warm Canvas (Talking Head)
-
-## Style card
-Name: Warm Canvas
-Family: Talking Head
-Look: A calm, editorial creator edit built around your own footage: full-frame talking head with a few clean serif captions, then a warm off-white canvas with a soft swoosh where white cards explain the idea while you shrink into a rounded picture-in-picture. Fraunces headlines in ink and sky blue, quiet Inter UI text, pastel chips.
-How it explains: Two modes and nothing else. On the A-roll, one or two captions fade up word by word with the key words in blue and small side chips slide in. On the canvas, a title chip lands, then one idea per beat: two cards with a dashed arrow, a comparison table whose numbers count up, a planner window where the bad blocks get flagged and removed, and three takeaway pills. Chapter cards push in vertically.
-
-## The look
-- Built for my own footage (A-roll): the speaker is the star and the graphics only explain. Keep the "Load A-roll" button and window.loadARoll(url) working so I can drop in my clip and see it in sync; until then the defocused stand-in shows where I sit in frame. If I give you my recording, time everything to it.
-- Two modes only. Mode 1: full-frame A-roll with at most one centered caption (white Fraunces 600 with a soft shadow, key words in sky blue #4f9cf0) and small white side chips with an icon and a two-line label. Mode 2: the warm canvas (#f7f5f4 with one big soft swoosh #ebe6e2), the speaker in a rounded 272x384 picture-in-picture at the right with a thin blue ring, and the content centered in the left column (COLX).
-- Type: Fraunces 600 for headlines, captions and card titles, tracked -2%, in two tones (ink #111111 plus sky blue for the key phrase). Inter 500 to 600 for UI labels, tags and small text in #2b2826 and #8f8a87. Title case for headlines.
-- Cards: white, radius 20, a very soft warm shadow, pastel tinted tags (blue, purple, green, yellow tints) and simple code-drawn icons. Never more than one window or two cards on the canvas at once.
-- One idea per beat: a title chip at the top of the column, then the explanation (cards with a dashed arrow, a table, a mock app window, or three takeaway pills). Each beat clears before the next arrives.
-- Motion is gentle: fades with a small rise (enter), soft counters, cards that slide together, never a bounce or a shake. Chapter cards (a blue number tile, a two-tone title and a progress dash row) push in and out vertically. Everything else is a clean cut.
-- Show me 2 to 4 beats on the canvas between A-roll moments, and come back to the full-frame A-roll for the personal lines.
-
-## Rules that keep it in style
-- Change the story, the words and the props; do not change the medium. Draw every subject accurately for my topic and double-check any facts and numbers you show.
-- Every frame is a pure function of t. Use CM.rng(seed) and CM.noise1(seed) for all randomness, never Math.random. Caching static layers is fine.
-- It will be watched on phones: at 1280x720 keep headlines and chapter titles at 52 px or more, captions at 60 px or more, card titles at 28 px or more, and UI labels at 14 px or more. Fill the frame.
-- Keep each frame fast, well under 50 ms.
-- Give my topic its own demo content; do not reuse the reference's story ("Can AI really plan my week?") unless my content asks for it.
-
-## Sound
-- Quiet and tactile, built to sit under a real voice: soft paper slides (pink noise through a rising bandpass) for chips, cards and title chips, low wooden taps for list items, pills and table rows, a light keyboard patter for counters, a breathy swell for each vertical push, and one quiet pad under the canvas section.
-- No chimes, bells, pings, sparkles or risers, and no music beat. Never more than 2 sounds at once.
-- Full mix (no voiceover): no music bed beyond the one quiet pad under the canvas, at most 2 sounds at once, and plenty of air between them.
-- Sync is everything: the sound score reads the same TL times as the picture.
-
-## Helpers to reuse (from the original prompt)
-One self-contained HTML file (an artifact if you can), built on the reference file below:
-- Keep the page, the player (play/pause, scrub, time, sound toggle, loop, Load voiceover) and the CM and SFX blocks unchanged.
-- Replace the demo's story: TL, CUTS, PUNCH (the jump-cut zooms inside the A-roll), the chapter card titles, the caption words, the side chips, the scene functions, SCENES and scoreSound(). Keep and reuse the style's helpers (the palette C, SERIF and SANS, PIP and COLX, toneText(), shadowCard(), tag(), enter(), fadeOut(), the icon set, canvasBG(), the A-roll stand-in and fullARoll(), caption(), titleChip(), sideChip(), chapterCard(), the card, table and planner drawing patterns, the vertical push compositor, and the Load A-roll button with window.loadARoll()).
-- Aim for 15 to 40 seconds unless my content needs more, and set the duration in TL.D.
-- Keep window.seek(t), window.DURATION, window.scoreSound and window.SFX exposed, so the video can be rendered frame by frame.
-- Load the style's fonts before the first draw, as the reference does: Fraunces (500 to 700) and Inter (500 to 700).

package/.agent/skills/faceless-shorts/README.md

@@ -1,34 +0,0 @@
-# Faceless Shorts: a free Claude skill
-
-Narrated faceless shorts (YouTube Shorts, TikTok, Reels) made 100% in code on your own computer: script, AI voice,
-word-synced captions, animation, sound effects and music, rendered to MP4. No AI video, no API keys, no credits.
-
-**Styles** (each has a style sheet, a style guide and a finished example video project)
-1. **Deep Dive 3D** (flagship): Zack D. Films-like 3D "What happens if…?" explainers. Example: *What happens if you hold your breath?*
-2. **Night File 3D**: unsolved mysteries and true crime in dark 3D rooms. Example: *The man who jumped out of a plane* (D.B. Cooper).
-3. **Inside Out**: flat 2D x-ray explainers. Example: *What if you swallow gum?*
-4. **Stick Story**: stick-figure survival steps. Example: *What if the power went out for 3 days?*
-5. **Paper Atlas**: history in cut paper and maps. Example: *How one bakery burned down London*.
-6. **Tall Tale 3D**: funny true stories, flat 2D cartoon people in a sunny 3D world. Example: *The girl who was mailed to grandma*.
-
-**How it works**
-script.json → Kokoro voice (local, free) → Whisper word timings (local, free) → shots drawn in code, each cut and caption
-placed on its word → procedural sound + music → headless Chrome captures every frame → ffmpeg renders the MP4.
-Claude checks a contact sheet of the whole video before you ever see it.
-Optional: when you ask, Claude uploads the finished short to your YouTube channel through Zapier (it confirms title, visibility and category first).
-
-**Needs:** Claude Code (or Claude with a linked computer), Node.js 22+, ffmpeg, Google Chrome, Python 3.10+, ~1 GB for the
-voice and Whisper models. Claude installs what's missing.
-
-**Windows:** run the scripts in Git Bash (installed with Git for Windows; Claude Code uses it already).
-
-**Install:** put this folder at `~/.claude/skills/faceless-shorts/`, then run `bash scripts/setup.sh` (prints "All set")
-and `bash scripts/selftest.sh` (prints "SELFTEST PASSED"). Start a new session and ask:
-"Make a Deep Dive 3D short: what happens if you crack your knuckles?"
-
-**Layout:** `SKILL.md` (what Claude follows) · `scripts/` (setup, selftest, new_video, voice, words, estimate) ·
-`styles/<style>/` (`style-sheet.png`, `STYLE.md`, `example/` = a complete project that renders as-is).
-
-**Licenses:** the code, stories, drawings and sound are original and free to use in your videos. Inter font: SIL Open Font
-License 1.1. Installed on first use: Kokoro-82M (Apache-2.0) via kokoro-onnx (MIT), faster-whisper + Whisper small.en (MIT),
-numpy, scipy.

package/.agent/skills/faceless-shorts/SKILL.md

@@ -1,124 +0,0 @@
----
-name: faceless-shorts
-description: Make narrated faceless YouTube Shorts / TikToks / Reels fully in code on the user's computer, with a free local AI voice and word-synced captions, $0. Six styles - Deep Dive 3D (Zack D. Films-like 3D "what happens if" body/science explainers), Night File 3D (unsolved mysteries, true crime), Inside Out (flat 2D x-ray explainers), Stick Story (stick-figure "what would you do if" survival steps), Paper Atlas (history in cut paper with maps), Tall Tale 3D (funny true stories with flat 2D cartoon people in a sunny 3D world, one-word captions). Use whenever someone wants a faceless short, a what-if or what-happens-if explainer, a Zack D. Films style video, a mystery or true-crime short, a stick-figure story, a history short, a funny true story or storytime cartoon short, or a video in one of these styles, even from just a topic. Can also upload the finished short to the user's YouTube channel through Zapier when they ask.
----
-
-# Faceless Shorts
-
-You make a finished vertical short from a topic: script, voice, word-synced captions, animated shots, sound and
-music, rendered to MP4. **Everything is code**: canvas drawing, headless Chrome capture, ffmpeg, a local Kokoro voice,
-local Whisper word timing, numpy sound. No AI images or video, no API keys, no credits.
-`SKILL` = the folder this file is in. Projects live in `~/faceless-shorts/<short-name>/`.
-
-| style | what it makes | length · aspect |
-|---|---|---|
-| **Deep Dive 3D** `deep-dive` | the flagship, Zack D. Films-like: sculpted 3D people in sunlit sets, glowing x-ray bodies, camera journeys through blood vessels, shallow focus. "What happens if…?" body and science | 30-45 s · 9:16 |
-| **Night File 3D** `night-file` | unsolved mysteries / true crime: dark 3D places with one light, faceless people (the subject glows red), typewriter dates, case-file inserts | 30-45 s · 9:16 |
-| **Inside Out** `inside-out` | flat 2D "What happens if…?" explainer: a simple hero, x-ray cutaways, bold 1-3 word captions, punchy cuts | 25-40 s · 9:16 |
-| **Stick Story** `stick-story` | "What would you do if…?" survival steps: stick people with acting eyes, flat rooms day/night, gold step cards | 30-45 s · 9:16 |
-| **Paper Atlas** `paper-atlas` | history in cut paper: parchment maps, crimson routes, gold date ribbons, period figures | 30-45 s · 9:16 |
-| **Tall Tale 3D** `tall-tale` | funny true stories: flat 2D cartoon people (big round heads, ink outlines) acting in a sunny 3D small-town world, camera always moving, one-word captions | 40-60 s · 9:16 |
-
-Which style: body / health / science "what happens if" → **Deep Dive 3D** (or **Inside Out** for a lighter, faster 2D
-look) · creepy, unexplained, true crime, "this happened in 1971" → **Night File 3D** · survival, life hacks, "here's what
-you do" → **Stick Story** · a historical event, "how X happened" → **Paper Atlas** · a funny or clever true story, a con, a loophole, "this guy did X" → **Tall Tale 3D**. If it isn't obvious, ask once.
-
-Each style = `styles/<style>/style-sheet.png` (the look in one image, drawn by the style's own code) + `STYLE.md` (the rules + the engine manual) +
-`example/` (a complete project that renders as-is). **Read the style's STYLE.md in full before planning.**
-
-## 0. First use on a computer
-`bash SKILL/scripts/setup.sh` until it prints **All set** (checks Node 22+, ffmpeg, Chrome; builds `~/.faceless-shorts/venv`;
-downloads the Kokoro voice ~340 MB and Whisper small.en ~0.5 GB). Install anything missing the free, official way, and ask
-before anything that needs a password. Then `bash SKILL/scripts/selftest.sh` must print **SELFTEST PASSED**: open the
-contact sheet and clip it names and look at them.
-
-## 1. Intake (ONE message, defaults filled in)
-Ask only what's unknown, all at once: topic (or "you pick 3 ideas"), a link to a short whose format they like (optional),
-length (default ~30 s), voice (`SKILL/scripts/py SKILL/scripts/voice.py --list`; default `am_michael`), voiceover mode:
-**AI voice** (default) · **their own recording** (they record after seeing the script; `words.py --audio`) ·
-**captions-only** (no voice yet; `estimate.py`), music (built-in bed · none · their own royalty-free file), end line / handle.
-"Just go" = the defaults.
-
-## 2. The plan (ONE screen, then wait)
-Show: title · the full narration, one line per cut, with its caption chunks (`WHAT IF *YOU | SWALLOW *GUM?`) · the shot list,
-one line each (what we see, what moves, which word it lands on) · new components you'll add to the kit, if any · facts and
-their sources. Ask **"Build it?"** and wait.
-
-**Script rules.** Write it the way a person actually talks to a friend: full, flowing sentences with connecting words
-("Well,", "where", "but", "so", "which means"), not a list of short caption-like statements. ~90-110 words for 35 s.
-**A line is a CUT, not a sentence**: split each sentence into 1-3 lines at natural pauses (commas, "and", "where"), so
-the picture changes every 1.5-3.5 s while the voice keeps flowing (voice.py speaks each whole sentence in one pass).
-Shape: hook question, then escalation, a myth or surprise, a twist ("but if you…"), a punchy last line that loops.
-Write numbers as they're said ("seven"). Read it out loud in your head: if it sounds like captions, rewrite it.
-Original writing only: take a format you like, never another creator's script or shots. Every fact is checked against a
-reliable source (list them in `script.json` "sources"). Health topics: accurate, no gore, no scary misinformation.
-
-## 3. Build
-1. `bash SKILL/scripts/new_video.sh <style> <short-name>` copies the example into `~/faceless-shorts/<short-name>/`.
-2. Write `script.json` (lines + caps + gaps; see STYLE.md).
-3. Timing: `SKILL/scripts/py SKILL/scripts/voice.py <proj>` then `SKILL/scripts/py SKILL/scripts/words.py <proj>`
-   (own recording: `words.py <proj> --audio their.m4a`; captions-only: `estimate.py <proj>`).
-4. Rewrite `src/shots.js`: one shot per line (or two for long lines), each placed with `cut(i)` and its key moments on
-   words with `Wd('word', line)`. Reuse the kit (`src/core/kit.js`); add new components there, in the style's look.
-5. In the project: `node tools/capture.mjs check` (must say CHECK OK) → `node tools/capture.mjs cues` →
-   `SKILL/scripts/py tools/mix.py .` → QA (section 4) → `node tools/capture.mjs video renders/<name>.mp4`.
-   Iterating on one shot: `node tools/capture.mjs frames review/f "4.2,4.6"` and look.
-
-## 4. QA: never show a frame you haven't looked at
-`node tools/capture.mjs sheet review/all.jpg 0.5 10` (a frame every 0.5 s, time-stamped) and read the whole sheet, then key
-frames full size. For every frame ask:
-- **Does it make sense?** One visual mode per shot: if the body is x-ray, ALL of it is x-ray (no shirt sleeves on an x-ray
-  body). Props move with whatever holds them (gum stays in the hand until the hand reaches the mouth). Nothing appears before
-  its cause.
-- **Does it match the narration at that moment?** The picture changes just before its line; reveals land on their word.
-- **Is anything colliding?** Moving things never cross labels, pills, the caption zone (y 1250-1550) or each other by accident.
-- **Captions:** 1-3 words, exactly one yellow word, spelled right, on the word being said, readable on a phone.
-- **Style sheet:** only the palette, flat shapes, x-ray views in cyan, danger in red.
-- **Sound:** voice clear over the music; effects soft (no harsh ticks or beeps).
-Fix, re-render the sheet, re-check. Tell the user what you caught and fixed.
-
-## 5. Deliver
-Open the MP4 for the user. Also give: a share copy if > 35 MB (`ffmpeg -i in.mp4 -c:v libx264 -crf 21 -preset slow -c:a aac
--b:a 160k -movflags +faststart share.mp4`), the best cover frame (`capture.mjs frames`), a title, a caption and 5-8 hashtags.
-For own-recording mode, also give them the script with each line's timing to read against. Offer one round of precise changes
-and re-render only what changed.
-
-## 5b. Publish to YouTube via Zapier (only when the user asks to upload)
-All calls are Code Mode tools: run them inside `execute` as `tools.zapier.*` (do not call them as normal tools). Validated once in a real upload; the values below are exact.
-
-1. **Connect.** If `list_zapier_skills` has a saved skill for uploading, use it. Otherwise make sure the YouTube app is enabled:
-   `discover_zapier_actions({ app: "YouTube" })` → `enable_zapier_action({ selected_api: "YouTubeV4CLIAPI", app_display_name: "YouTube", action: "upload_video" })`.
-   If no connection exists yet, `manage_zapier_connections({ selected_api: "YouTubeV4CLIAPI" })` returns a connect URL for the user.
-2. **Host the file.** The action's `video` / `thumbnail` params need a URL Zapier's servers can fetch; a local path is silently wrong.
-   Put the MP4 on a temp host, in this order (all tested 2026-11-03; on Windows use `curl.exe`):
-   - a. **litterbox** (1 h to 72 h expiry, up to 1 GB): `curl -F "reqtype=fileupload" -F "time=1h" -F "fileToUpload=@renders/<name>.mp4" https://litterbox.catbox.moe/resources/internals/api.php` → returns the direct URL as plain text.
-   - b. **uguu.se** (about 3 h expiry, good for shorts under 35 MB): `curl -F "files[]=@renders/<name>.mp4" https://uguu.se/upload` → returns JSON; use `files[0].url` (the `d.uguu.se/...` link).
-   - c. **catbox** (permanent, no expiry): same call as (a) without `-F "time=1h"`. Public forever, so only with the user's OK.
-   - Skip these (verified broken 2026-11-03, don't waste retries): 0x0.st (uploads disabled), tmpfiles.org (link is a viewer page, not the raw file), gofile.io (raw download needs an account token), file.io (curl API redirects to its web app), bashupload.com, temp.sh (dead).
-   - **Always verify** the URL before using it: `curl -sSI <url>` must return 200/206 with a video content-type. Never trust a host's JSON without checking.
-   - Tell the user this puts their video on a public temp host that expires by itself.
-3. **Confirm first.** This writes to the user's real channel, so confirm before executing: title, visibility (private / unlisted / public) and category.
-   Resolve `category_id` with `inspect_zapier_actions({ tool_name: "youtube_upload_video", enum_property: "category_id" })` if needed.
-   Comedy = 23 (default suggestion for Tall Tale and Deep Dive), Education = 27, Pets & Animals = 15.
-4. **Upload.**
-   `execute_zapier_write_action({ selected_api: "YouTubeV4CLIAPI", action: "upload_video", tool_name: "youtube_upload_video", params: { video: "<temp URL>", title, description (end it with #shorts), tags: [5-8 hashtags without #], category_id: "23", privacy_status, made_for_kids: false, embeddable: true, notify_subscribers: true, default_language: "en", license: "youtube" } })`
-5. **Report.** The result contains `id`; the link is `https://www.youtube.com/watch?v=<id>`. A vertical video of 60 s or less becomes a Short automatically (no extra field).
-   Open the local file for the user (`tools.browser.preview({ path: "renders/<name>.mp4" })`) and give the YouTube URL in text. The in-app browser often fails to load youtube.com with ERR_ABORTED; that is not a failed upload.
-6. **Thumbnails.** Unlisted uploads don't take a custom thumbnail through this action; offer the saved cover frame if the user later makes the video public.
-
-## Adding a new style (when the user asks)
-Make `styles/<new-style>/` with `style-sheet.png`, `STYLE.md` (same sections as Inside Out: look rules, palette, format,
-engine/kit map, components, shot anatomy, cue types, QA pitfalls) and a complete `example/`. Copy Inside Out's
-`tools/`, `index.html`, `src/main.js`, `src/core/{lib,timing,captions}.js`, then build the new look's kit and shots. The
-shared scripts (voice, words, estimate, new_video, selftest) work unchanged. Add a row to the table above and run
-`selftest.sh` until every style passes.
-
-## 3D styles (Deep Dive, Night File, Tall Tale)
-They render real 3D with three.js in headless Chrome. With a graphics card it's fast and full quality. Without one
-(a laptop without GPU acceleration, a cloud sandbox) render with `SOFTWARE_GL=1 QS='&res=0.6&aa=0&shadow=1024'
-FPS=24` in front of `node tools/capture.mjs ...` (≈2 s per frame; tell the user the render time up front, run it in the
-background, and QA with `sheet review/all.jpg 1 10` first). Their STYLE.md files list the 3D API.
-
-## Never
-Generate images or video with AI; spend the user's money or credits; upload to YouTube or any host without the user asking and confirming; copy another creator's script, shots or frames;
-overwrite a user's project without asking; ship a frame you haven't looked at.

package/.agent/skills/faceless-shorts/styles/deep-dive/STYLE.md

@@ -1,58 +0,0 @@
-# Deep Dive 3D: stylized 3D "What happens if…?" explainers, 9:16 (the flagship)
-
-Real 3D, rendered in the browser with three.js: sculpted stylized people in warm sunlit sets, glowing x-ray cutaways,
-camera journeys through blood vessels, shallow focus, a camera that never stops moving. Curious, shocking, satisfying.
-Example: `example/` = **WHAT HAPPENS IF YOU HOLD YOUR BREATH?** (37 s, 11 shots). Style sheet: `style-sheet.png`
-(regenerate it by opening `example/sheet.html` through `node tools/capture.mjs preview`).
-
-## 1. The look (rules)
-- **Three worlds, and every video moves between them:** (A) the real world: a person in a warm, sunlit set (kitchen),
-  (B) the x-ray: the same body as cyan glass with a glowing skeleton and organs, (C) inside the body: macro shots
-  (blood vessel, organs) in deep red with bloom. Real world → x-ray → inside → back out for the ending.
-- **Warm low sun outside, cyan inside the body.** X-ray cyan `#4FD1E8` only for x-ray glass; red `#B8283A` for blood.
-- **Shallow focus always:** focus on the subject, soft background (`aperture` 0.003-0.006 in the real world).
-- **The camera always moves:** every shot pushes in, orbits or drifts (lerp the camera between two points over the shot).
-- **No gore.** Injury or danger is shown as x-ray, glow, red alarm colour or a warning sign, never wounds.
-- People: one sculpted body, many looks (`new Human({ skin, shirt, pants, shoes, hair, iris })`). Faces are calm and
-  readable; avoid hand close-ups (hands are simple mittens).
-- Captions: sentence case, 1-3 words, white with a dark outline, one yellow word (`src/core/captions.js`, y 1480).
-  Labels: small dark pills with a leader line (`label()`), counters/facts as pills (`pill()`), `warning()` for danger.
-
-## 2. Format
-- 1080x1920, 24 fps (cinematic, and faster to render), 30-45 s. Narration `am_michael` x1.0, conversational.
-- One line = one shot (cut just before the line: `cut(i)`), and a long line can hold a second shot starting on a word
-  (`Wd('blood', 2)`), e.g. x-ray lungs → cut on "blood" → inside the vessel.
-- Music: `"tension"` (low drone + slow pulse) for body/danger topics, `"bed"` for light ones.
-
-## 3. Engine map (`example/src/`)
-| file | gives you |
-|---|---|
-| `core/engine.js` | `makeEngine(canvas, { w, h, samples })` → `E.render(scene, cam, look)`: ACES tone, soft shadows, room environment light, depth of field, bloom, vignette + grain. `look = { focus, aperture, maxblur, bloom, bloomThreshold, exposure, vignette, grain, warm }`. `lightRig(scene, {...})`, `camera(fov)` |
-| `core/sdf.js` | sculpting: `sphere, ellipsoid, capsule, rbox, torus, smin/smax/ssub, noise3`, `mesh(sdf, bounds, step, colorFn)` (surface nets → smooth BufferGeometry with vertex colours). Use it for any new organ, prop or set piece so everything shares the soft clay look |
-| `core/human.js` | `new Human(look)`: `.group` (feet at y 0, faces +z), `.pose({ spine, head, shL, elL, shR, elR, hipL, knL, hipR, knR, lid, look })` (degrees; lid -0.4 wide … 1 closed), `.walk(phase, amt)`, `.setBodyMaterial(m)` (x-ray), `.parts[0]` = torso (scale it to breathe), `.spine` (attach insides), `.hips.position.y` (sit: 0.66 + hips 88°, knees 88°) |
-| `core/sets.js` | `kitchen(E)`, `xrayStage(E)`, `vessel(E)` (`.update(t, camZ, co2, dark)`), `xrayMaterial(color)`, `insides()` → `{ group, stomach, heart, lungs }` |
-| `core/look2d.js` | 2D layer over the 3D: `caption`, `label(text, x, y, tx, ty, scale)`, `pill(text, x, y, scale, colour)`, `warning(x, y, s)`, `blackout(k)` |
-| `src/shots.js` | the per-video file. `screen(vec3, cam)` gives the 2D point of a 3D spot (aim labels at organs) |
-| `src/main.js` | renders the 3D frame into the hidden `#gl` canvas, copies it to `#cv`, then 2D + captions |
-
-## 4. A shot
-```js
-add({ id: 'S03', at: cut(2), note: 'x-ray chest, lungs full', draw(lt, d, t, blit) {
-  reset(); lungs(1);                                                     // pose everything as a pure function of time
-  cam.position.copy(lerpV(V3(0.55, 1.5, 1.6), V3(0.3, 1.42, 1.1), eio(lt / d))); cam.lookAt(0, 1.33, 0);
-  E.render(X.scene, cam, { ...XLOOK, focus: cam.position.distanceTo(V3(0, 1.33, 0)) }); blit();
-  label('LUNGS: FULL', 760, 520, ...screen(V3(0.1, 1.42, 0.05), cam), eob(seg(lt, 0.2, 0.45)));
-} });
-```
-Keep 2D pills and labels off faces (y 1150-1250 over the torso, or top y 300 when the head is low in frame).
-
-## 5. Rendering speed
-3D frames are heavy without a graphics card. On a normal computer with a GPU it renders at full quality. On a machine
-without one (or a cloud sandbox) render with `SOFTWARE_GL=1 QS='&res=0.6&aa=0&shadow=1024' FPS=24` (≈2 s per frame,
-~30 min for a 37 s short) and check with `sheet ... 1` (one frame per second) before the full render.
-
-## 6. QA pitfalls we've hit
-- A camera inside a light's glow cone washes the frame flat (cones are volumetric meshes): keep cameras out of them.
-- Seat backs / props between camera and subject: keep the camera in clear space (centre of an aisle).
-- Overlays on faces: check every frame where a pill or warning appears.
-- The x-ray body must be x-ray everywhere except the head (`setBodyMaterial`).

package/.agent/skills/faceless-shorts/styles/deep-dive/example/data/cues.json

@@ -1,140 +0,0 @@
-{
- "duration": 37.383,
- "cues": [
-  {
-   "t": 0.12,
-   "type": "inhale",
-   "dur": 1
-  },
-  {
-   "t": 3.704,
-   "type": "whoosh"
-  },
-  {
-   "t": 3.754,
-   "type": "pop",
-   "gain": 0.6
-  },
-  {
-   "t": 6.228,
-   "type": "whoosh"
-  },
-  {
-   "t": 6.478,
-   "type": "pop",
-   "gain": 0.5
-  },
-  {
-   "t": 8.19,
-   "type": "whoosh"
-  },
-  {
-   "t": 8.19,
-   "type": "slide",
-   "dur": 2.7539999999999996,
-   "gain": 0.5
-  },
-  {
-   "t": 10.944,
-   "type": "whoosh"
-  },
-  {
-   "t": 11.044,
-   "type": "bubbles",
-   "dur": 2.8900000000000006
-  },
-  {
-   "t": 13.834,
-   "type": "whoosh"
-  },
-  {
-   "t": 14.034,
-   "type": "alarm",
-   "dur": 4.478000000000001
-  },
-  {
-   "t": 18.512,
-   "type": "whoosh"
-  },
-  {
-   "t": 18.762,
-   "type": "thud",
-   "gain": 0.45
-  },
-  {
-   "t": 19.382,
-   "type": "thud",
-   "gain": 0.45
-  },
-  {
-   "t": 20.002,
-   "type": "thud",
-   "gain": 0.45
-  },
-  {
-   "t": 20.622,
-   "type": "thud",
-   "gain": 0.45
-  },
-  {
-   "t": 21.242,
-   "type": "thud",
-   "gain": 0.45
-  },
-  {
-   "t": 21.862,
-   "type": "thud",
-   "gain": 0.45
-  },
-  {
-   "t": 22.272,
-   "type": "whoosh"
-  },
-  {
-   "t": 22.422,
-   "type": "heartbeat"
-  },
-  {
-   "t": 22.922,
-   "type": "heartbeat"
-  },
-  {
-   "t": 23.512,
-   "type": "heartbeat"
-  },
-  {
-   "t": 24.208,
-   "type": "heartbeat"
-  },
-  {
-   "t": 25.03,
-   "type": "heartbeat"
-  },
-  {
-   "t": 28.233,
-   "type": "slide",
-   "dur": 1.2,
-   "gain": 0.5
-  },
-  {
-   "t": 31.491,
-   "type": "inhale",
-   "dur": 1
-  },
-  {
-   "t": 33.091,
-   "type": "inhale",
-   "dur": 1,
-   "gain": 0.7
-  },
-  {
-   "t": 33.156,
-   "type": "whoosh"
-  },
-  {
-   "t": 33.406,
-   "type": "slam",
-   "gain": 0.7
-  }
- ]
-}

package/.agent/skills/faceless-shorts/styles/deep-dive/example/data/lines.json

@@ -1,67 +0,0 @@
-{
- "duration": 37.383,
- "voice": "am_michael",
- "speed": 1.0,
- "lines": [
-  {
-   "i": 0,
-   "text": "What happens if you hold your breath for as long as you can?",
-   "start": 0.35,
-   "end": 3.524
-  },
-  {
-   "i": 1,
-   "text": "For the first thirty seconds, nothing much.",
-   "start": 3.784,
-   "end": 6.048
-  },
-  {
-   "i": 2,
-   "text": "Your lungs are full, and your blood is still carrying plenty of oxygen.",
-   "start": 6.308,
-   "end": 10.764
-  },
-  {
-   "i": 3,
-   "text": "But carbon dioxide starts building up,",
-   "start": 11.024,
-   "end": 13.394
-  },
-  {
-   "i": 4,
-   "text": "and that, not the lack of oxygen, is what makes you desperate to breathe.",
-   "start": 13.914,
-   "end": 18.332
-  },
-  {
-   "i": 5,
-   "text": "Your diaphragm starts twitching, trying to force a breath,",
-   "start": 18.592,
-   "end": 21.872
-  },
-  {
-   "i": 6,
-   "text": "and your heart actually slows down to save oxygen.",
-   "start": 22.352,
-   "end": 25.568
-  },
-  {
-   "i": 7,
-   "text": "Push even further, and your brain simply switches you off.",
-   "start": 25.828,
-   "end": 29.208
-  },
-  {
-   "i": 8,
-   "text": "You pass out, and your body starts breathing again on its own.",
-   "start": 29.468,
-   "end": 32.976
-  },
-  {
-   "i": 9,
-   "text": "Which is why you should never try this underwater.",
-   "start": 33.236,
-   "end": 36.183
-  }
- ]
-}