Skills clear-prompt

Clear Prompt Engineering

install

source · Clone the upstream repo

git clone https://github.com/openclaw/skills

Claude Code · Install into ~/.claude/skills/

T=$(mktemp -d) && git clone --depth=1 https://github.com/openclaw/skills "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/371166758-qq/clear-prompt" ~/.claude/skills/clawdbot-skills-clear-prompt && rm -rf "$T"

manifest: skills/371166758-qq/clear-prompt/SKILL.md

source content

Clear Prompt Engineering

A skill that transforms vague prompts into precise, actionable instructions using 5 core principles.

When to Use

Use this skill when the user's message is ambiguous, too broad, or lacks necessary context. Apply it proactively before responding to help users craft better prompts.

5 Core Principles

明确任务目标: Say what to do directly, no ambiguity
补充场景背景: Provide usage context and prerequisites
明确输出要求: Specify format, length, and purpose
提供关键信息: Attach necessary data and details
排除无关选项: State what NOT to do (reduces ambiguity)

Template

Structure every prompt as:

【场景背景】+【已有信息】+【任务目标】+【输出要求】

Response Behavior

When you receive a vague prompt, instead of guessing:

Identify which principles are missing
Provide a before/after example showing the improvement
Suggest a refined version the user can confirm

Quick Templates

Config tasks

"I'm on 【OS】, already have 【credential】, please give me CLI commands to 【action】, ready to copy-paste"

Publish tasks

"I want to publish 【skill type】 to 【platform/org】, named 【name】, give me complete steps + prerequisites"

Debugging

"Running 【command】 got error 【error message】, already have 【conditions】, excluding 【known issues】, help troubleshoot"

Writing

"Write 【doc type】, highlighting 【key strengths】, must be 【style】, under 【length】 words"

Usage

This skill activates automatically when detecting vague input. No manual trigger needed.