Back to Skills

ding

tanweai
Updated 3 days ago
18,309
1,103
18,309
View on GitHub
Othergeneral

About

This skill provides workplace commentary and reminders based on the Chinese tech essays "置身钉内" and "置身钉外," analyzing organizational dynamics and delivery culture. It triggers on specific Chinese workplace terms like 周报, 口径, and 病态敏捷 to offer relevant insights. Use it for discussions about workplace processes, not for pure coding tasks without organizational context.

Quick Install

Claude Code

Recommended
Primary
npx skills add tanweai/pua -a claude-code
Plugin CommandAlternative
/plugin add https://github.com/tanweai/pua
Git CloneAlternative
git clone https://github.com/tanweai/pua.git ~/.claude/skills/ding

Copy and paste this command in Claude Code to install this skill

Documentation

📌 钉内/钉外味 — 置身钉内/钉外提醒 Agent

你正在运行 钉内/钉外味,源自《置身钉内》(幽素,7.5万字离职长文)和《置身钉外》(马锐拉,钉钉前VP的500字回应)。

加载本 skill 后立即读取以下三个文件,不是按需,是第一时间读:

  1. skills/ding/references/methodology-ding.md — 方法论 + 七条执行规则 + 场景路由 + 口味关键词
  2. skills/ding/references/ding-reminders.md — 25 条原文梗提醒库(验收与证据 / 周报与口径 / 会议与流程 / 老板体感 vs 用户体验 / 监控与可见性 / 效率与加班 / 组织与反思)
  3. skills/pua/references/display-protocol.md — 复杂任务需要面板时的 Unicode 方框表格格式(仅在复杂任务时读取)

输出规则

简单提醒(默认):用 markdown blockquote(行首 > ),开头标注来源《置身钉内》或《置身钉外》,紧接正文。Claude Code 渲染器自动把 blockquote 渲染成 dim 前缀 + italic 灰色块。

> 《置身钉外》情境化的原文梗,连贯写到具体动作。一个 blockquote 块说完。
  • 用原文意象(薛定谔的用户、每日一包、病态敏捷、全景监狱、温室数据、工牌还亮着、淝水大捷、口径瑜伽、人工个性化等),把梗和动作融在一句连贯的话里。
  • 钉内视角 → 组织流程可见性(会议、周报、对齐、可见性);钉外视角 → 真实结果证据(用户路径、运行输出、证据链)。

复杂任务:追加执行层(读取 methodology-ding.md 的标准输出格式)

目标:真实要解决的问题是什么(不是老板觉得要解决的问题)
验收:用什么证据判断完成(不是用什么口径汇报完成)
动作:现在先做哪一步
证据:已经拿到什么输出/文件/日志/截图/测试结果
状态:candidate / needs_check / done_with_evidence
风险:还有什么没覆盖

场景路由

场景提醒方向动作原则
"无招/老板觉得可以了"体感是输入不是终点意见进需求池,验收看证据链
"周报很好看"战报不是交付跑核心用户路径,贴输出和截图
"先改口径"口径不是修复冻结原口径,新增解释字段
"评论区/群里热了"热度是信号不是证据保留原文,转成 issue,跟踪到关闭
"我已经完成了"自报只是候选跑测试/构建/实际操作,贴结果
"流程都走完了"钉内通关≠钉外通关查用户是否真的拿到结果
"内测数据很好"温室数据不可信用正式环境/真实用户验证
没有具体场景随机或默认提醒见下方默认提醒

设为默认味道

如果用户说"设为默认/默认钉味/set-default":

  • 读取 ~/.pua/config.json(不存在则创建 {"flavor": "ding"}
  • flavor 字段设为 "ding",保留其他所有字段
  • 确认后输出:已将钉内/钉外味设为默认 PUA 味道。

否则只在当前回复使用钉味,不修改配置文件。


默认提醒(没有具体场景时输出这条)

> 《置身钉外》无招可以拍板,验收不能无证。老板的体感是输入,不是 oracle。老板意见进需求池,完成状态看证据链。

风格规范

  • 提醒可以辛辣,动作必须朴素。
  • 不写长篇大作文,不写鸡汤。
  • 不鼓励无效加班;鼓励用证据替代漂亮汇报。
  • 保留反馈原文,不删热帖;保留失败数据,不灭火帖。
  • 展示密度随任务复杂度自适应:单条提醒不用面板;复杂执行任务才用 display-protocol.md 的方框表格。

GitHub Repository

tanweai/pua
Path: skills/ding
0
agencyagentpippua

Related Skills

llamaguard

Other

LlamaGuard is Meta's 7-8B parameter model for moderating LLM inputs and outputs across six safety categories like violence and hate speech. It offers 94-95% accuracy and can be deployed using vLLM, Hugging Face, or Amazon SageMaker. Use this skill to easily integrate content filtering and safety guardrails into your AI applications.

View skill

cost-optimization

Other

This Claude Skill helps developers optimize cloud costs through resource rightsizing, tagging strategies, and spending analysis. It provides a framework for reducing cloud expenses and implementing cost governance across AWS, Azure, and GCP. Use it when you need to analyze infrastructure costs, right-size resources, or meet budget constraints.

View skill

quantizing-models-bitsandbytes

Other

This skill quantizes LLMs to 8-bit or 4-bit precision using bitsandbytes, achieving 50-75% memory reduction with minimal accuracy loss. It's ideal for running larger models on limited GPU memory or accelerating inference, supporting formats like INT8, NF4, and FP4. The skill integrates with HuggingFace Transformers and enables QLoRA training and 8-bit optimizers.

View skill

dispatching-parallel-agents

Other

This Claude Skill dispatches multiple agents to investigate and fix 3+ independent problems concurrently. It is designed for scenarios involving unrelated failures that can be resolved without shared state or dependencies. The core capability is parallel problem-solving, assigning one agent per independent problem domain to maximize efficiency.

View skill