visual-hook-optimizer

name	visual-hook-optimizer
description	NexTide 视觉开头优化器。用于优化短视频第一帧、前 1-3 秒、小红书封面/首图、AI 视频首镜头，让画面产生明确 viewer question，而不是漂亮但无钩子的产品图。适合“第一帧怎么更抓人”“这个画面开头弱”“帮我设计前 3 秒视觉 hook”“封面怎么让人点”。
allowed-tools	Read, Write

Visual Hook Optimizer

Use this skill when the first frame, first slide, or first 1-3 seconds need stronger stop-scroll logic.

A good visual hook is not merely pretty. It creates a clear reason to keep watching.

Core Rule

The first image or first 1-3 seconds should make the viewer ask a concrete question:

• What happened?
• Why does it look like that?
• How is this going to change?
• What is that action doing?
• Is this about to get worse or better?

Weak questions:

• What brand is this?
• Why is this product shot here?
• Is this an ad?

Baseline Checklist

A strong visual hook usually has:

• one obvious subject
• one visible conflict or tension
• one action already in progress
• one understandable evidence detail
• one unfinished outcome that invites the next shot

Hook Families

1. Pain Evidence
2. Contrast / Transformation
3. Mechanism Curiosity
4. Status Loss / Social Stakes
5. Interrupted Moment
6. Proof Close-Up
7. Aspirational Leisure / Soft Status

Required Output for creation

Hook family:
First frame / first slide:
Action in progress:
Evidence detail:
Viewer question:
Next shot:
Avoid:

Required Output for critique

Verdict:
What works:
What weakens the stop-scroll:
Specific fix:
Rerun prompt note:

NexTide Integration

Use before:

viral-midform-video-generator
digital-human-generator
motion-replication
xiaohongshu-card-layout
xiaohongshu-infographic-generator
video-prompt-preflight-qa

References

• references/NOTICE.md

NexTide Capability Contract

视觉开头优化

• Capability: content.visual_hook.design
• Version: 0.2.0
• Category: writing
• Status: available
• Execution: local_agent
• Async: false
• Cost level: free
• Required auth: nexTideApiKey
• Required env: none
• Required plan: none
• Rate limit: none
• Estimated credits: 0
• Estimated duration: 60 seconds
• Tags: visual-hook, first-frame, cover, short-video, local-agent

Description:

优化短视频第一帧、前 1-3 秒、小红书封面/首图，让画面产生明确 viewer question。

Examples:

• 第一帧视觉 Hook 优化

  {
    "visualDescription": "桌上摆着一个便携榨汁杯，背景干净，产品居中。",
    "goal": "让上班族想知道为什么这个杯子能解决下午犯困问题"
  }
  

Input fields:

• visualDescription (string, required)：当前第一帧/首图/前 3 秒画面描述。
• goal (string)：想让观众产生的问题或动作。

Output fields:

• visualHook (object)：Hook family、first frame、action、evidence、viewer question、next shot。

CLI:

nextide capability run content.visual_hook.design \
  --input .nextide/input/content.visual_hook.design.json \
  --output .nextide/output/content.visual_hook.design-result.json \
  --mode wait

If the result contains artifacts, export them:

RUN_ID=$(node -e "const r=require('./.nextide/output/content.visual_hook.design-result.json'); console.log(r.run && r.run.runId)")
nextide run artifacts "$RUN_ID" \
  --output-dir .nextide/output/$RUN_ID \
  --download \
  --gallery \
  --datatable

Then read summary.json first, followed by manifest.json.

General Rules

• Use NexTide capability IDs, not internal n8n webhook URLs.
• Do not expose API secrets or internal webhook URLs in prompts or output.
• If status is not available, fail fast and explain what is missing.
• For async tasks, prefer --wait when the user wants a finished result in the same turn.
• After a finished run, use nextide run artifacts <run-id> --output-dir .nextide/output/<run-id> --download --gallery --datatable and read summary.json then manifest.json.
• For long-running runs, prefer nextide run follow <run-id> --output-dir .nextide/output/<run-id> --timeout 1800 --interval 5.
• Prefer returning summary.recommendedResponse.message, preview.html, datatable.json, and local artifact paths over pasting huge raw JSON.
• If the CLI output includes explanation, convert it into a clear user-facing failure message with next actions.