prompt-polish - Rewrite rough prompts into model-specific optimized prompts.
Rewrites rough prompts into model-specific prompts based on official Anthropic guides, supporting Claude Fable 5 and Claude Opus 4.8.
Tags
Updated: 2026-09-18Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Parse model and prompt invocations
- Diagnose raw prompts across four axes
- Rewrite prompts using official model doctrine
- Validate polished prompts against hard gates
Inputs
- Raw prompt draft
- Target model name or alias
- Model reference documentation
- Praxis Workflow OS contract
Outputs
- Polished prompt in code block
- List of bracketed placeholders
- Optional execution note
- Clarification questions
Requirements
- Support for Claude Fable 5 or Opus 4.8
- Model reference documentation files