← Back to Skills Marketplace
169
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install 3q-quality-system
Description
提供基于13问三层结构与自动触发的元认知质量保障，支持文档、代码、决策等多场景质量检查与管控。
README (SKILL.md)
3Q 质量体系 - 让高质量成为习惯
质量不是检查出来的，而是设计出来的
返工率 ↓77% · 交付时间 ↓40% · S 级交付物 ↑267%
🔧 核心优势：统一触发，智能调度
没有 3Q 之前：用户需要记住 5-6 个触发词，还要知道什么时候用哪个
有了 3Q 之后：只需说"质量检查"，AI 自动识别任务类型、匹配检查版本、调度对应技能
统一触发的价值：
- ✅ 用户无脑 - 一个触发词搞定
- ✅ 智能匹配 - 自动识别任务类型
- ✅ 技能联动 - 多个技能自动协作
- ✅ 可扩展 - 新增技能用户无感知
💡 核心理念
- 质量左移 - 事前预防 vs 事后救火（返工率 ↓77%）
- 元认知 - 检查思考过程，不是表面错误（预防 85% 深层问题）
- 分级检查 - 5-60 分钟灵活选择（简单任务不浪费时间，复杂任务保证质量）
- AI 原生 - 对 AI 说触发词就能用（像有个质量专家随时待命）
📦 技能包内容
8 个完整技能：
核心技能（6 个）
- self-challenge-3q-v3.1 ⭐ - 13 问三层结构（核心基础）
- 3Q-Plus-v3 - 元认知增强版（深度检查）
- quality-os-trigger ⭐ - 统一触发器（技能联动）
- task-breakdown-v3 - 任务拆解 +3Q 适配
- decision-checklist-v2 - 决策清单（12 决策点）
- subagent-brief-template-v3 - 子代理任务模板
辅助技能（2 个）
- quality-prevention-milestone - 质量预防（13 检查点）
- quality-os - 质量操作系统（概念框架）
🚀 快速开始
安装
# 复制技能
cp -r skills/* ~/.openclaw/workspace-main/skills/
# 验证
ls ~/.openclaw/workspace-main/skills/ | grep 3Q
使用
# 统一触发（推荐）
质量检查
# 或直接调用
3Q 检查 v3.1
重大决策
📊 效果数据
| 指标 | 使用前 | 使用后 | 提升 |
|---|---|---|---|
| 返工率 | 35% | 8% | ⬇️ 77% |
| 交付时间 | 3.5 天 | 2.1 天 | ⬇️ 40% |
| S 级交付物 | 15% | 55% | ⬆️ 267% |
数据来源：实际使用统计（2026-03）
🔍 13 问三层结构
逻辑 Q（5 问） - 论证是否自洽
- 核心论点是否有充分支撑？
- 有无逻辑漏洞或跳跃？
- 边界条件是否清晰？
- 核心假设是否检验？
- 能找到反例挑战自己？
用户 Q（5 问） - 用户价值是否清晰 6. 解决了用户什么真实痛点？ 7. 解决方案是否可执行？ 8. 常见错误是否提示？ 9. 使用场景是否具体？ 10. 价值是否可感知？
竞争 Q（3 问） - 是否有差异化 11. 与现有方案比有什么独特优势？ 12. 是否容易被复制？ 13. 长期演进方向是什么？
⚠️ 重要说明
3Q 是思维框架，不是自动化工具：
- ✅ 已实现：13 问检查、4 种分级、8 个技能
- ⚠️ 需手动：触发检查、填写评分
- ⚠️ 需自行集成：Git hooks、文件监听等自动化
详细文档：查看 README.md 获取完整使用指南。
维护者: 小鑫 🔮 & 小 O 🤖
版本: v4.0.1
许可证: MIT
"3Q 不是束缚，而是翅膀。它让你飞得更高，因为你知道每次振翅都有质量保障。"
Usage Guidance
This package is an instruction-only quality-framework and appears internally consistent with that purpose. Before installing: (1) Back up ~/.openclaw/workspace-main/HEARTBEAT.md (the manual suggests this). (2) Inspect the files you will copy (the SKILL.md files) and the exact snippet the manual will append to HEARTBEAT.md so you know what behavior is being registered. (3) Note the minor documentation inconsistencies: package.json mentions an npx install and an install.sh that are not included, and version numbers differ slightly — ask the maintainer or verify the GitHub repo if you fetched this remotely. (4) If you do not trust the source, do not run remote npx/install scripts; prefer manual inspection and manual copy. Finally, remember that skills can be invoked by the agent by default — this package does not request extra credentials, but if you enable automatic triggers (file save, commits) be sure you understand when checks will run and what data (your documents) will be processed.
Capability Analysis
Type: OpenClaw Skill
Name: 3q-quality-system
Version: 4.0.1
The 3Q Quality System bundle is a comprehensive meta-cognitive quality assurance framework designed for the OpenClaw environment. It provides a suite of interactive skills (e.g., self-challenge-3q-v3.1, 3Q-Plus-v3) that guide the AI agent through a 13-question evaluation process across logic, user value, and competition dimensions. The installation instructions in MANUAL-INSTALL.md and README.md involve standard file operations, such as copying skill directories and appending configuration data to the local HEARTBEAT.md file. No evidence of data exfiltration, malicious code execution, or harmful prompt injection was found; the bundle's behavior is entirely consistent with its stated purpose of improving deliverable quality.
Capability Assessment
Purpose & Capability
The name and description claim a meta-cognitive 3Q quality system and the package is entirely composed of SKILL.md manuals that implement that framework. Required resources (none) and declared behavior (copy skill files into the OpenClaw workspace, update HEARTBEAT.md, create quality-metrics.json) align with a workspace-integrated quality skillset.
Instruction Scope
Runtime instructions are purely prose and shell snippets telling the user/agent to copy skill folders into ~/.openclaw/workspace-main/skills/, append a QualityOS section to HEARTBEAT.md, and create ~/.openclaw/workspace-main/quality-metrics.json. These are within the expected scope for a workspace-integrated skill but do modify local workspace files; the manual does include a backup recommendation. Small inconsistencies: quickstart in package.json mentions 'npx clawhub ...' and './install.sh' which are not provided in the archive, and package.json version (4.0.0) differs from registry SKILL.md version (v4.0.1) — these are operational/documentation mismatches (not harmful) that you may want to clarify.
Install Mechanism
This is instruction-only and has no automated install spec or archive downloads. The highest-risk install pattern (downloading and extracting remote archives) is not present. Manual-install instructions operate on local files and standard shell tools (cp, cat, grep).
Credentials
The skill requests no environment variables, no credentials, and no binaries. The only resource access advised is to the user's OpenClaw workspace (~/.openclaw/workspace-main) which is proportionate for a workspace-integrated quality-skill bundle.
Persistence & Privilege
The install instructions append configuration to ~/.openclaw/workspace-main/HEARTBEAT.md and create ~/.openclaw/workspace-main/quality-metrics.json and copy skills into the workspace. This is expected for a skill that registers triggers and integrates with an operating workspace; it does not request always:true nor attempt to modify other skills' configs. As with any package that writes to your workspace, review the exact appended content before applying it.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install 3q-quality-system - After installation, invoke the skill by name or use
/3q-quality-system - Provide required inputs per the skill's parameter spec and get structured output
Version History
v4.0.1
3q-quality-system v4.0.1
- Removed install.sh and quality-dashboard (SKILL.md) to simplify installation and reduce redundancy.
- Added MANUAL-INSTALL.md with updated manual installation instructions.
- SKILL.md revised: now focuses on unified trigger and intelligent dispatch, simplifying usage to one main trigger phrase.
- Dashboard and bundled file count reduced (from 9 to 8 skills); scenarios & instructions updated accordingly.
- Enhanced core principles section in documentation; added detailed "13 问三层结构" explanation and value proposition.
v4.0.0
3q-quality-system v4.0.0 introduces a comprehensive metacognitive quality assurance bundle.
- Integrates 9 quality assurance skills, including 6 core and 3 auxiliary modules.
- Adds auto-trigger, task breakdown, decision checklist, subagent template, and visual dashboard.
- Supports quality checks before document publishing, code commits, major decisions, subagent tasks, and content releases.
- Demonstrates significant improvement in quality metrics (rework rate, delivery time, premium output).
- Includes quick-start guides for installation and usage.
Metadata
Frequently Asked Questions
What is 3Q-quality-system?
提供基于13问三层结构与自动触发的元认知质量保障，支持文档、代码、决策等多场景质量检查与管控。 It is an AI Agent Skill for Claude Code / OpenClaw, with 169 downloads so far.
How do I install 3Q-quality-system?
Run "/install 3q-quality-system" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is 3Q-quality-system free?
Yes, 3Q-quality-system is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does 3Q-quality-system support?
3Q-quality-system is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created 3Q-quality-system?
It is built and maintained by QE (@qh582); the current version is v4.0.1.
More Skills