Diagnose and rewrite LLM prompts with explicit output formats and boundary constraints.
npx skills add https://github.com/CNife/skills --skill prompt-craft-cnife
Please help me install this Agent Skill. Skill: prompt-craft Source: https://github.com/CNife/skills/tree/main/.archive/prompt-craft Command: npx skills add https://github.com/CNife/skills --skill prompt-craft-cnife
This Skill helps users revise an existing LLM prompt when the model's output is vague, unstable, or poorly structured. It keeps the original task intent while making the instruction clearer and easier to execute.
把你要优化的提示词直接发给我，我会先给出诊断，再提供可直接复制使用的改进版。
High-intent search queries and answers about installing and using this skill.
Refine existing LLM prompts to improve clarity and controllability by applying structured diagnosis, instruction prioritization, delimiter-based context separation, and explicit output format constraints while preserving the original language and intent.
Rewrite prompts for extraction or summarization by applying structured diagnosis and instruction prioritization, adding explicit output formatting and boundary constraints to ensure the LLM produces stable and controllable results.
The best way to add output formatting and boundary constraints to LLM prompts is through delimiter-based context separation and instruction prioritization, ensuring the LLM produces structured and stable results.
Yes, you can optimize prompts for response drafting while preserving the original language and style, using instruction design and diagnosis to improve prompt clarity and output stability without altering the original intent.
Your LLM prompt produces poorly structured results due to ambiguity, missing constraints, and weak formatting, which can be resolved through structured prompt diagnosis and adding explicit output format and boundary constraints.
Delimiter-based context separation is recommended for instruction design to clearly separate context and improve prompt clarity, ensuring the LLM understands boundaries and produces stable, well-structured output.
calvingitAudit and rewrite LLM prompts to improve clarity and alignment with user intent.
fratilanicoDesign APEX OS agent prompts with CoT/ToT patterns and JSON-schema constraints.
patricio0312revDiagnose LLM output failures including hallucinations, format violations, and reasoning errors.
mcauduro0Analyze and optimize LLM prompts with Python scripts.
