跳到正文
cccskills
免费在 GitHub 公开

pair

TDD pair programming with mode switching. Claude can be driver (writes code) or navigator (watches, questions, provokes). Supports GitHub issues, file watching, or arbitrary prompts. Scientific TDD, baby steps, one test at a time. Trigger on: pair, let's pair, tdd, you drive, I'll drive, dojo.

中文简介正在准备中。当前显示原文说明。

集成工具GitHub
查看安装方法

包含的文件(1)

  • SKILL.md5.0 KB

SKILL.md(原文)

安装前可以查看交给 AI 代理的指令内容。

Pair — TDD Pair Programming

Two modes, one skill. Switch anytime.

  • Driver mode — Claude writes code, user navigates
  • Navigator mode — User writes code, Claude watches and questions
  • Vibe mode — Claude goes autonomous until done

Usage

  • /pair <issue_url> --driver — Claude drives on a GitHub issue
  • /pair <issue_url> --navigator — Claude navigates on a GitHub issue
  • /pair <issue_number> — GitHub issue, ask who drives
  • /pair <file_or_dir> — Watch files (navigator mode implied)
  • /pair <prompt> — Arbitrary problem, ask who drives
  • /pair — Ask what we're building and who drives

Switching Modes

The user can switch at any time:

  • "you drive" / "drive" / "switch" -> Claude becomes driver
  • "I'll drive" / "my turn" / "switch" -> Claude becomes navigator
  • "vibe" / "just do it" -> Claude goes autonomous
  • "stop" / "back" -> Return to previous pairing mode

Phase 1: Understand Together (both modes)

Before any code, set the stage.

  1. Fetch context

    GitHub: gh issue view <number> --json title,body Prompt: Restate the problem back to confirm understanding.

  2. Explore the codebase — read relevant files. Explain what exists, how things work, where new behavior fits. Trace data flow from CLI -> APIServer -> domain -> Store/ETS.

  3. Expand the problem — edge cases, constraints, implications.

  4. Propose initial scenarios — lean list of behaviors (3-5 max):

"I see these behaviors:

  1. ...
  2. ...
  3. ...

Where should we start? Missing something?"

  1. Wait for feedback. Discuss until aligned.

Driver Mode

Claude writes code. User thinks, questions, directs.

The Loop

Checkpoint 1: Which test?

Propose the next test. Explain what behavior it captures. Ask.

Wait. The navigator may redirect, refine, question.

Checkpoint 2: RED

Write the test. Run it with mix test test/path:LINE. Show the failure.

Wait. Share thinking for the green. Ask the navigator's take.

Checkpoint 3: GREEN

Discuss approach. Explain what you'd write and why. Ask.

Wait. Write the minimum code agreed on. Run mix test. Show GREEN.

Checkpoint 4: REFACTOR

Propose cleanup if warranted. If not, say so and move on.

Wait. Refactor only what's approved. Run mix test.

REPEAT

Back to Checkpoint 1. Update behavior list as needed.

Driver Rules

  • Never advance without navigator input
  • Explain what you're doing and why
  • One test at a time
  • The navigator can ask you to do structural work (rename files, refactor, move things around)

Navigator Mode

User writes code. Claude watches, questions, provokes thinking.

File Watcher (optional)

If a file or directory is provided, start the watcher:

  • File: fswatch -1 <file>
  • Directory: fswatch -1 -r <dir>

On trigger: read, run mix test, report RED/GREEN, restart watcher.

Output

GREEN: GREEN. {one-line summary} RED: RED. {what failed and why}

Navigator Behavior

Problem before solution. Always.

  • Be critical. Provoke the driver to think. Don't hand out answers
  • Ask questions: "What do you expect this to return?" "What's the simplest case?" "What if the input is nil?"
  • When stuck, ask a question that unblocks thinking. Don't give a snippet
  • Challenge assumptions: "Do we need this yet?" "Is that the right abstraction?"
  • Only give code when explicitly asked
  • Point out bugs as questions: "Is that the right pattern match?"

Navigator Rules

  • Never write code unless explicitly asked
  • Don't suggest next steps unless asked
  • Don't explain what the code does (the driver wrote it)
  • Don't recap what changed
  • No filler

Vibe Mode

When triggered ("vibe", "just do it"):

  1. Delegate to /dev if the project's dev skill exists (.claude/skills/dev/). The project's dev skill has project-specific agents, review gates, and conventions.

  2. If no /dev skill exists, fall back to autonomous TDD:

    • Assess done vs remaining
    • Build a concrete plan
    • Present plan and wait for approval
    • Execute following TDD (RED-GREEN-REFACTOR)
    • Run mix test && mix dialyzer && mix smoke
    • Report back

User can interrupt vibe mode anytime to return to pairing.


Shared Rules (all modes)

  1. Scientific TDD — failing test FIRST, verify RED, minimal fix, verify GREEN, revert fix to verify RED again
  2. Problem before solution — expand the problem, then attack it
  3. One test at a time — no batching, no skipping
  4. Baby steps — if it feels big, split it
  5. No filler — discuss the problem, not feelings
  6. Run tests after every change — mix test
  7. Lean — think small, think now, divide to conquer

评价

还没有评价。欢迎分享你的使用感受。

评论

无需登录即可发表。欢迎分享使用感受、提问,或说明在哪些环境下无法运行。

还没有评论。来写第一条吧。

提交后会立即公开。请勿填写邮箱、电话等个人信息。URL 最多 2 个。

0 / 1000 字

此表单受 reCAPTCHA 保护,适用 Google 的隐私权政策和服务条款。

同一仓库的技能

bugfix

免费

Bug hunter. Reproduces bugs with failing tests (RED), then fixes with TDD. Accepts a prompt, issue URL, or bug description. Use when: bugfix, fix bug, debug, broken, regression, failing, doesn't work, fix this.

中文简介正在准备中。当前显示原文说明。

leandronsp/overmind122026年4月13日 更新

commit

免费

Create a git commit following project conventions. Use when: commit, commit this, make a commit, commit changes, git commit, save changes, commit my work, stage and commit, detailed commit.

中文简介正在准备中。当前显示原文说明。

leandronsp/overmind122026年4月13日 更新

debug

免费

Elixir debugging workflow and tools. Use when: debug, debugging, pry, inspect, trace, why is this failing, what's wrong, investigate, diagnose.

中文简介正在准备中。当前显示原文说明。

leandronsp/overmind122026年4月13日 更新

dev

免费

Implementer - fetches a GitHub issue, builds an implementation plan, then implements using TDD. Use when: implement, build this, code this, add this feature, TDD, test first, red green refactor, pick a task, next task, dev.

中文简介正在准备中。当前显示原文说明。

leandronsp/overmind122026年4月13日 更新

learn

免费

Extract learnings from the current session and update memory/config. Use when: learn from this, what did we learn, update memory, session retrospective, extract patterns, save learnings.

中文简介正在准备中。当前显示原文说明。

leandronsp/overmind122026年4月13日 更新

overmind

免费

Orchestrator - spawns an autonomous orchestrator agent that decomposes a task into subtasks and drives multiple Claude agents to completion. Use when: overmind, orchestrate, multi-agent, delegate, spawn agents, break down task, run agents.

中文简介正在准备中。当前显示原文说明。

leandronsp/overmind122026年4月13日 更新

查看 leandronsp 的全部技能

报告此技能的问题