skills.rest
webdevarif
webdevarifCOMMUNITY
Explore all skills by this author →
Website𝕏@webdevarifGitHub

prompt

Refines rough prompts into structured expert-level prompts using project CLAUDE.md context.

Updated Apr 18, 2026
Search Tags:#prompt engineering#ai workflow#claude.md#project context#structured prompts#prompt refinement
One-click install
npx skills add https://github.com/webdevarif/claude-skills --skill prompt-webdevarif
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: prompt
Source: https://github.com/webdevarif/claude-skills/tree/main/prompt-refiner
Command: npx skills add https://github.com/webdevarif/claude-skills --skill prompt-webdevarif

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Vague, messy, or shorthand prompts produce inconsistent AI results because they lack project context, constraints, and clear structure. This Skill transforms raw input into polished, structured prompts grounded in the project's own CLAUDE.md rules. ## Core Features & Use Cases - Project-Aware Refinement: Reads CLAUDE.md first to inject correct component names, import paths, data patterns, and styling rules into the refined prompt. - Structured Prompt Template: Outputs prompts with Role, Task, Project Context, Constraints, Expected Output, and Suggested Skill sections. - Multilingual Input Handling: Accepts rough input in English, Bangla, mixed language, or shorthand and preserves the original intent. - Use Case: A developer types a vague request like "add login page with validation" and receives a complete prompt referencing the project's actual form components, routing patterns, and type locations defined in CLAUDE.md. ## Quick Start Invoke the /prompt command with your rough request, such as asking it to refine a vague feature description into a structured prompt for your project.

Frequently Asked Questions about prompt

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I turn a vague prompt into a structured prompt?▼

Invoke the /prompt command with your rough input in any language or shorthand. The Skill reads your project's CLAUDE.md, identifies your core intent, and outputs a structured prompt with role, task, context, constraints, and expected output sections.

How does prompt refinement use CLAUDE.md project rules?▼

The Skill reads CLAUDE.md from the project root before refining anything, extracting tech stack, component rules, file organization, and banned patterns. It then injects only the rules relevant to your specific task into the refined prompt.

Can I refine prompts without a CLAUDE.md file?▼

Yes, the Skill falls back to generic mode when no CLAUDE.md exists. It still restructures your rough input into a clear prompt with role, task, constraints, and expected output, but without project-specific context like component names or file paths.

Does prompt refinement work with non-English input?▼

Yes, the Skill accepts input in Bangla, English, mixed language, or shorthand. It identifies the core intent regardless of phrasing and preserves what you want while improving how the request is structured.

What are the limitations of project-aware prompt refinement?▼

The quality of project context depends entirely on how complete and accurate the CLAUDE.md file is. If CLAUDE.md is missing, outdated, or vague, the refined prompt will lack specific file paths, component names, and pattern references.

Related Skills

Explore More Software Engineering Skills→
tdiminotdimino
41

claude-md-manager

Creates and maintains CLAUDE.md documentation files for Claude Code projects.

Community
Intermediate
mattwoodcomattwoodco
3

add-feature

Route project and feature development workflows to specialized child skills.

Community
Intermediate
opmauopmau
6

review

Review staged git diffs against CLAUDE.md rules before commit.

Community
Intermediate
rahlplxrahlplx

clean-code

Enforce pragmatic coding standards like SRP, DRY, KISS, and YAGNI.

Community
Basic