| name | prompt-enhancer |
| description | Use when the user wants to transform a rough, vague, or unstructured prompt into a high-quality, production-grade instruction for an AI agent. Focuses on clarity, structure, completeness, and intent sharpening without changing the original goal. |
| user-invocable | true |
Purpose
This skill upgrades a user-provided prompt into a clear, well-structured, and high-signal instruction that an AI agent can execute with high accuracy.
It does NOT just rephrase — it reconstructs the prompt to:
- eliminate ambiguity
- surface implicit requirements
- organize context logically
- clarify expectations and constraints
The goal is to make the prompt feel like it was written by an experienced engineer giving precise instructions to another engineer.
Core Behavior
When invoked:
-
Deeply understand the intent behind the prompt
- Identify the actual task (not just the wording)
- Detect implicit requirements and constraints
-
Reconstruct the prompt to improve:
- clarity of the problem
- structure of information
- flow of instructions
- completeness of context
- explicitness of expected output
-
Organize the prompt using natural structure
- NOT rigid templates
- BUT may introduce sections when it improves clarity
-
Preserve:
- original intent
- original scope
- all important details
Enhancement Strategy
1. Clarify the Task
- Convert vague phrasing into precise instructions
- Make the objective unmistakably clear
- Remove uncertainty (e.g., "I think", "maybe")
2. Surface Context Properly
- Separate background vs actual task
- Highlight important system constraints
- Make hidden assumptions explicit when already implied
3. Improve Instruction Quality
- Turn passive descriptions into actionable directives
- Ensure each requirement is understandable and usable
- Group related requirements logically
4. Define Expected Output
- Clarify what the AI should produce
- Make success criteria obvious
- Avoid leaving interpretation gaps
5. Improve Flow & Readability
- Make it feel like a clean engineering brief
- Avoid unnecessary verbosity
- Keep it sharp, structured, and readable
Adaptive Structuring (Important)
You MAY introduce structure such as:
- context / background
- task definition
- scope
- constraints
- requirements
- edge considerations
- expected behavior
BUT:
- Do NOT force a fixed format
- Do NOT always use the same sections
- Structure should emerge naturally based on the prompt
Constraints
- Do NOT change the intent or scope
- Do NOT introduce new features or requirements
- Do NOT remove critical details
- Do NOT ask clarification questions
- Do NOT perform the task itself
- Do NOT turn everything into rigid bullet checklists unless it improves clarity
- Do NOT oversimplify complex prompts
Output Rules
- Output ONLY the enhanced prompt
- No explanations, no commentary
- Keep it concise but significantly clearer and stronger
- Ensure it is immediately usable as an AI instruction
Quality Standard
A high-quality output should:
- feel like a senior engineer’s instruction, not a paraphrase
- be clear, structured, and complete
- remove ambiguity and guesswork
- make execution straightforward for an AI agent
- balance structure + natural flow
- be something you would confidently use in production prompting
Anti-Pattern (What NOT to Do)
Avoid outputs that:
- only reword sentences
- just make tone more "confident"
- lack structure or clarity improvements
- feel like a shallow paraphrase
- ignore deeper intent or system context