Rewrite rough ideas into precise, actionable prompts for AI execution.
npx skills add https://github.com/mslshao/claude-harness --skill refine-mslshao
Please help me install this Agent Skill. Skill: refine Source: https://github.com/mslshao/claude-harness/tree/main/dotclaude/skills/refine Command: npx skills add https://github.com/mslshao/claude-harness --skill refine-mslshao
This Skill helps when you have a vague, incomplete, or overly terse prompt and need to turn it into something an AI can execute reliably without guessing. It reduces back-and-forth by converting rough intent into a clear, scoped, actionable prompt.
Ask the refine skill to rewrite your rough task description into a complete prompt for Claude Code with the right context, assumptions, and scope boundaries.
High-intent search queries and answers about installing and using this skill.
To turn a rough idea into a precise prompt, you can use prompt refinement to extract user intent and infer missing context from your workspace. This rewrites incomplete instructions into actionable prompts with explicit goals, scope, and constraints for reliable AI execution.
The best way to scope a broad feature idea is to apply task scoping during prompt refinement. This flags overly broad tasks and suggests a tighter framing, converting vague intent into a focused prompt with explicit boundaries and constraints.
Context gathering works by pulling missing project context, such as file paths and current git state, from your conversation and workspace state. This information is inferred and integrated into the refined prompt to ensure the AI has the necessary background.
Yes, you can use prompt refinement to tighten an existing prompt for refactoring, debugging, or investigation work. It sharpens your initial request by rewriting it with explicit goals and assumptions to reduce back-and-forth guessing.
The AI guesses because terse instructions lack explicit goals, scope, and context. Prompt refinement solves this by extracting your actual intent, inferring missing workspace information, and producing a complete prompt with stated assumptions.
Yes, prompt refinement works without specific dependencies or components. It functions as an intermediate-level skill that processes your conversation and workspace state directly to produce actionable, project-aware prompts.
HermeticOrmusCapture development session state into a HANDOFF.md file.
danhannah94Generate a deterministic Claude Code execution prompt from a refined CSDLC ticket.
musher-devGenerate project-level Claude Code agents with configured tools, models, and permissions.
ankitjh4Diagnose prompt weaknesses and rewrite prompts into structured Claude 4.x instructions.
