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。
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。
安装前可以查看交给 AI 代理的指令内容。
Two modes, one skill. Switch anytime.
/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 drivesThe user can switch at any time:
Before any code, set the stage.
Fetch context
GitHub: gh issue view <number> --json title,body
Prompt: Restate the problem back to confirm understanding.
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.
Expand the problem — edge cases, constraints, implications.
Propose initial scenarios — lean list of behaviors (3-5 max):
"I see these behaviors:
- ...
- ...
- ...
Where should we start? Missing something?"
Claude writes code. User thinks, questions, directs.
Propose the next test. Explain what behavior it captures. Ask.
Wait. The navigator may redirect, refine, question.
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.
Discuss approach. Explain what you'd write and why. Ask.
Wait. Write the minimum code agreed on. Run mix test. Show GREEN.
Propose cleanup if warranted. If not, say so and move on.
Wait. Refactor only what's approved. Run mix test.
Back to Checkpoint 1. Update behavior list as needed.
User writes code. Claude watches, questions, provokes thinking.
If a file or directory is provided, start the watcher:
fswatch -1 <file>fswatch -1 -r <dir>On trigger: read, run mix test, report RED/GREEN, restart watcher.
GREEN: GREEN. {one-line summary}
RED: RED. {what failed and why}
Problem before solution. Always.
When triggered ("vibe", "just do it"):
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.
If no /dev skill exists, fall back to autonomous TDD:
mix test && mix dialyzer && mix smokeUser can interrupt vibe mode anytime to return to pairing.
mix test还没有评价。欢迎分享你的使用感受。
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。
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。
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。
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。
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。
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。
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。