one-detail-reads-person-video-gen

# One detail reads a person (cue → example → debate line) **Template (timings assume `duration` 10; rescale for 5 or 15):** **0–3s** *you can read someone by ONE cue*; **3–8s** **one concrete example** (user-chosen: e.g. how they treat service staff, how they handle a small promise, phone posture at dinner—**no** real non-public figures); **8–10s** **provocative summary** that invites disagreement (e.g. *THAT’S THE TELL—FIGHT ME* toned **playfully**, not harassment). **English subtitles** with **explicit timing**. Default **`duration`** is **10** when unspecified. **Dependencies:** `WERYAI_API_KEY` + Node.js 18+. **`scripts/video_gen.js`** + **[`WERYAI_VIDEO_API.md`](resources/WERYAI_VIDEO_API.md)**. **No other Cursor skills required.** ## Prerequisites - `WERYAI_API_KEY` **must be set**; Node **18+**. Prefer public **`https`** image URLs; if the assembled `scripts/video_gen.js` supports local file paths, review/verify the script and explicitly consent before local read-and-upload to WeryAI. - Non-empty `model`. ## Security, secrets, and API hosts - **`WERYAI_API_KEY`**: Secret; **`requires.env`** / **`primaryEnv`**. - **API hosts** fixed in `video_gen.js`; no URL env overrides. - **Local image handling disclosure**: Prefer public **`https`** image URLs. If the assembled `scripts/video_gen.js` supports local file paths, it may read a local image and upload it to WeryAI to obtain a public URL; require review / verification and explicit consent before using that path. - **Higher assurance**: Prefer a short-lived or isolated environment; review `scripts/video_gen.js` before production use. Verify whether the runtime can read local image files and upload them to WeryAI, and obtain explicit consent before using that path. ## Prompt expansion (mandatory) - **Duration:** Default **10**; **5** or **15** when allowed and requested. - **Timeline scaling:** Rescale **0–3 / 3–8 / 8–10** proportionally; keep **thesis → example → closer** order. **Locked rhythm (example for duration 10):** - **0.0–3.0s** — Thesis: e.g. *ONE DETAIL EXPOSES THEIR LEVEL*. - **3.0–8.0s** — Example beat (split into **two** on-screen lines if needed, e.g. **3–5.5** and **5.5–8.0**) describing **generic silhouettes**, not identifiable people. - **8.0–10.0s** — Closer: **strong opinion**, **playful fight-me** energy—**no** slurs, **no** targeting protected groups. **Visuals:** Close-ups **hands, eyes, cafe table, door holds**; shallow DOF; **constant** micro-motion. **Typography:** Bold sans, outline, lower third; **English** on screen. **Negatives (Kling):** illegible text, watermark. **Confirmation:** Full `prompt` in table. ### Niche checklist - **Example** must be **filmable in B-roll**, not a lecture. - Avoid **diagnosing** real individuals or mental illness. ## Workflow 1. Text / image / multi-image (**SEEDANCE_2_0**, ≤3). 2. Collect **which cue** the user cares about. 3. Expand → validate → confirm → `wait`. 4. **`[Video](url)`** for results. ## CLI reference ~~~sh node scripts/video_gen.js wait --json '{"model":"KLING_V3_0_PRO","prompt":"…","duration":10,"aspect_ratio":"9:16","generate_audio":true,"negative_prompt":"illegible text, watermark"}' node scripts/video_gen.js wait --json '{"model":"SEEDANCE_2_0","prompt":"…","duration":10,"aspect_ratio":"9:16","resolution":"720p","generate_audio":true}' ~~~ **Full reference:** **[`WERYAI_VIDEO_API.md`](resources/WERYAI_VIDEO_API.md)**. ## Definition of done **[Video](url)** or error; timed **English** captions follow **thesis / example / closer** structure with **ranges scaled to `duration`** (or user-approved split). ## Boundaries (out of scope) - **No** harassment, hate, or doxxing framing. - Not professional psychology advice. - Captions via **prompt** only; CLI details in **`resources/WERYAI_VIDEO_API.md`**. - **Multi-image** only **SEEDANCE_2_0**, max **3**. ### Example prompts - `Cue: how they treat waiters—cafe B-roll, playful closer` - `Cue: punctuality on small promises—office hands clock` - `Image: anonymous crowd silhouette URL → same beat structure` - `Cue: how they apologize after a mistake—cafe reenactment, playful closer` --- ## Model and API constraints (frozen for this skill) ### Text-to-video | model_key | durations | aspect_ratios | resolutions | audio | negative_prompt | prompt limit | |-----------|-----------|---------------|-------------|-------|-----------------|--------------| | KLING_V3_0_PRO | 5, 10, 15 | 9:16, 1:1, 16:9 | *(omit)* | yes | yes | 2000 | | KLING_V3_0_STA | 5, 10, 15 | 9:16, 1:1, 16:9 | *(omit)* | yes | yes | 2000 | | SEEDANCE_2_0 | 5, 10, 15 | 9:16, 1:1, 16:9 | 480p, 720p | yes | no | 2000 | ### Image-to-video | model_key | durations | aspect_ratios | resolutions | audio | negative_prompt | image slots | |-----------|-----------|---------------|-------------|-------|-----------------|-------------| | KLING_V3_0_PRO | 5, 10, 15 | 9:16, 16:9, 1:1 | *(omit)* | yes | yes | single `image` | | KLING_V3_0_STA | 5, 10, 15 | 9:16, 16:9, 1:1 | *(omit)* | yes | yes | single `image` | | SEEDANCE_2_0 | 5, 10, 15 | 9:16, 1:1, 16:9 | 480p, 720p | yes | no | up to **3** `images` | ## Recommended models | Tier | model_key | Default duration | |------|-----------|------------------| | ⭐ Best | KLING_V3_0_PRO | 10 | | 👍 Good | KLING_V3_0_STA | 10 | | ⚡ Fast | SEEDANCE_2_0 | 10 (`resolution`: **720p**) | ## Default parameters | Field | Value | |-------|--------| | aspect_ratio | 9:16 | | duration | 10 | | generate_audio | true | | resolution | 720p (Seedance) | | negative_prompt | Kling: `illegible text, garbled captions, watermark` | ## Text-to-video (primary) Collect **cue** → map to **three windows** → confirm → `wait`. ## Image-to-video (optional) Reference still → emotional read with same structure. ## Tips - Closer works when it’s **obviously rhetorical**, not cruel. - **Translate** user Chinese brief to **English** captions unless they specify another on-screen language. > Packaged as **`one-detail-reads-person-video-gen`**.