
from higgsfield-ai-prompt-skill20
Compose production-ready video and image prompts for the Higgsfield platform using the MCSLA formula, named presets, and genre templates for cinematic results.
This skill provides a complete, production-aware prompt-writing workflow for Higgsfield — a multi-engine cinematic AI platform. It routes requests to focused sub-skills, enforces the MCSLA (Model · Camera · Subject · Look · Action) structure, appends shared negative constraints, and outputs ready-to-paste prompts tailored to named models, camera controls, and motion presets. It includes template-based fast paths for common genres and a full production path for Cinema Studio workflows.
Use this skill whenever a user asks to write or refine Higgsfield prompts, select models or motion presets, troubleshoot failed generations, or prepare Cinema Studio projects. Trigger on requests like “write me a video prompt,” “which model for action scenes,” or any direct mention of higgsfield.ai and platform-specific presets.
Inferable agents: Claude-style skills and assistants that can run stepwise skill reads and produce structured prompt output (Claude, Copilot-style agents).
Higgsfield is a comprehensive prompt-engineering skill for the Higgsfield AI video/image generation platform. It provides a structured MCSLA formula, routing tables to 20+ sub-skills, genre templates, and strict output formatting. No scripts were bundled or executed. The SKILL.md is exceptionally detailed with clear triggers, mandatory workflows, and hard rules to prevent stale knowledge usage.
Pure instructional skill with no executable code. Well-designed routing architecture with progressive disclosure. The mandatory file-read rule before generating prompts is a thoughtful guardrail against stale training data.