11 social hook
Model-aware Claude Agent Skills for AI image & video generation on Higgsfield — 15 creative skills with opt-in MCP generation. Reworked from beshuaxian/higgsfield-seedance2-jineng.
npx -y skills add pixelab-ch/higgsfield-skills --skill 11-social-hookAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Crafts viral social media hook video prompts for TikTok, Instagram Reels, and YouTube Shorts via Higgsfield. Routes to kling3_0 (primary) or grok_video (fallback). Use when the user wants scroll-stopping hooks, viral short-form video, attention-grabbing openers, TikTok content, Reels content, Shorts content, or any social-media-optimized video.
SKILL.md
5.8 KB, as published. Nobody here has run it
Higgsfield Social Hook Skill
What this skill does
Crafts scroll-stopping hook video prompts for TikTok, Instagram Reels, and YouTube Shorts and routes them to the best Higgsfield model. On request, it assembles the full parameter set, surfaces the credit cost, and — after explicit user confirmation — generates the asset via the opt-in generation flow.
Model routing
Primary and fallback video models
| kling3_0 (primary) | grok_video (fallback) | |
|---|---|---|
| Rationale | Wide duration range, mode tuning, supports start + end image | Fallback when primary unavailable; native audio |
| Mode | std, pro, 4k* | No tunable params |
| Sound | on, off | Native audio (always on) |
| Aspect ratios | 16:9, 9:16, 1:1 | 16:9, 9:16, 1:1 |
| Duration | 3–15 s (continuous range) | 1–15 s (continuous range) |
| Media roles | start_image, end_image | start_image only (max 1) |
*kling3_0 4k mode is a generation-mode enum, NOT a 4K output-resolution claim.
Do not describe it as 4K video or high-resolution output.
Routing rule: Use kling3_0 by default. Only fall back to grok_video when
(1) the primary model is unavailable, or (2) the user wants native audio generation
and, after being informed of the trade-offs (no tunable params, no end_image,
start_image max 1), explicitly accepts those constraints. Never switch silently.
MODEL-06 directive: If a parameter is rejected at generation time, call
higgsfield:models_explore with the target model name to re-verify the live schema.
Full parameter tables: references/model-specs.md.
Per-platform aspect ratio quick rule
Vertical 9:16 for TikTok / Reels / Shorts; 16:9 for standard YouTube and LinkedIn;
1:1 for Instagram feed square posts (kling3_0 and grok_video both support 1:1).
Full per-platform recommendation table: references/model-specs.md.
Prompt-building workflow
-
Gather intent — Confirm with the user: content category (comedy, transformation, product, story, etc.), platform (TikTok/Reels/Shorts), target aspect ratio, desired duration, and whether native audio is needed.
-
Select model — Apply the routing table above. Use
kling3_0by default; selectgrok_videoonly when a routing trigger applies. -
Build the hook prompt — Use the craft principles and the 25+ hook patterns from references/hook-craft.md:
- Identify the hook type (visual shock, curiosity gap, emotional trigger, pattern interrupt, or direct address) and the 2-second master template structure.
- Apply per-platform optimization from references/platforms.md.
- Reference worked prompts by category from references/examples.md.
-
Present for review — Show the assembled prompt, hook structure, model, aspect ratio, and duration to the user for review and refinement before any generation call.
Opt-in generation
Generation costs Higgsfield credits and requires explicit user confirmation before any generate call is made. This skill never auto-generates.
Full step-by-step flow (confirmation gate, balance/cost surface,
generate → poll → display):
../../shared/generation-flow.md
This skill's primary model: kling3_0
Media upload: Conditional — only when the user supplies a start_image or
end_image reference for image-to-video. This skill is NOT a mandatory-I2V skill;
T2V is the default path. When the user provides a reference image, run the
higgsfield:media_upload → higgsfield:media_confirm atomic pair before generating.
Tool signatures: ../../shared/mcp-tools.md
Reference materials
| File | Contents |
|---|---|
| references/model-specs.md | Verified parameter tables for kling3_0 and grok_video, per-platform aspect ratio and duration recommendations, routing rationale, and re-verify directive |
| references/hook-craft.md | Hook psychology (5 attention triggers), the full 25+ proven hook pattern encyclopedia, the 2-second master template, retention techniques, sound strategy, text overlay strategy, and common mistakes |
| references/platforms.md | Per-platform optimization (TikTok/Reels/Shorts), content-category playbooks (comedy, educational, satisfying, transformation, BTS, reaction, storytime, product, lifestyle), trending format templates, and analytics-driven iteration guidance |
| references/examples.md | 5 production-ready worked hook prompts (satisfying transformation, comedy skit, product reveal, emotional story, impossible visual) with full timing breakdowns and parameter settings |