Higgsfield Social Hook

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 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

1. 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.

2. Select model — Apply the routing table above. Use kling3_0 by default; select grok_video only when a routing trigger applies.

3. 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.

4. 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