First Frame Video Start · Open Skill 公開スキル
First Frame Video Start
First Frame Video Start は Digen 上の公開 Open Skill です。開くと、このワークフローが何を作るか分かります。
Custom
1 0 1
https://d7qlwi7tl2u27.cloudfront.net/skill-covers/prod/4284864/2026-10-21/5n47sqr3.jpg
# First Frame Video Start Deliver one **cinematic motion clip that begins on the user's uploaded frame**. The first frame must match the upload exactly: same subject, clothing, environment, and composition. From there the clip continues with **lifelike full-body motion** (fluid limbs, torso, and head), **secondary motion** in hair and fabric, a **steady follow camera** with cinematic depth of field, and **consistent lighting, texture, and color grading** across every frame. No morphing, warping, flicker, or distortion. **Upload path:** optional still polish, then **`ref2v`** with the uploaded image as the first frame and the locked motion brief. **Direct generate:** text-to-image **cinematic keyframe** (default invented subject and scene), then the same `ref2v` pipeline. **Upload is never required.** Prompt templates live in `references/`. No structured project state. **Language lock:** This file and all of `references/` are **English only**. Reply in the **user's language**. Every `generate_image`, `generate_video`, and `video_agent` `instruction` is **English**. Do not tell the user tool names, channel names, or dump full prompts. ## User-visible speech (hard) **Allowed lines only:** - Need input: one short note that they may **upload a starting-frame photo**, **describe the scene**, or **tap direct generate**—then pause **P-empty** or **P-brief**. - Starting generate: at most **one** short line (user language), then tool calls only—**no** `set_guidance` on that turn. - Success: one or two sentences that the **first-frame motion video** is ready—no URL, no prompt dump. Pause **P-done**. - Tool failure: brief honest note. Pause **P-fail**. - Safety block: one polite refusal. Pause **P-refuse**. If the user already asked to **generate**, tapped **direct generate**, or sent a one-shot create message, **do not** re-ask—read the prompt references, apply defaults, and generate. ## Pause = clickable buttons (hard) Any turn that **stops without a new shown final video**, and **every success turn**, must end with: 1. One or two short sentences (user language). 2. **`set_guidance`** in the **same turn**, with **at least two** clickable options. **No tool calls after `set_guidance`.** Do **not** call `set_guidance` in the same turn as `generate_image`, `generate_video`, or `recognize_image`. Each button is an object with `text` and `action` (not bare strings): - **`send_message`**: `content` is the full sentence sent when tapped. - **`fill_input`**: `content` is starter text the user can edit. Offer **2–4** options. Localize `text` and `content` to the user's language. No channel or tool names on buttons. Before every pause turn, `read_skill_file` → `references/guidance-buttons.md`. | Pause | When | Example intents (localize) | |---|---|---| | P-empty | No upload, no scene brief, no direct-generate tap, no last clip to refine | Direct generate motion clip · Upload starting frame · Describe my scene | | P-brief | Wants motion video but no subject cues and no upload | Direct generate default scene · fill_input custom scene · Upload photo instead | | P-refuse | Real-person face used without consent, gore, hate, NSFW, or harmful motion | Direct generate fictional subject · Upload a different photo | | P-fail | Generation or `video_agent` `success=false` after retry | Direct generate again · Regenerate from last still · Calmer camera, same motion | | P-done | Final first-frame motion video ready | Direct generate new scene · Upload my photo · Slower motion · Longer clip | ## When to use - User wants **First Frame Motion Video** / **animate this image** / **make my photo move** / **start from this frame** as a cinematic motion clip. - User uploads a **photo** and wants it to be the **exact first frame** while the subject continues with natural motion. - User has **no photo** or taps **direct generate**—run **T2I keyframe + ref2v** using defaults in `references/guidance-buttons.md` (default invented cinematic subject, not