Назад към всички

craftwork

// 9 engineering disciplines for AI agents — TDD, debugging, code review, planning, and multi-agent coordination. Build software like it matters.

$ git log --oneline --stat
stars:1,933
forks:367
updated:March 4, 2026
SKILL.mdreadonly
SKILL.md Frontmatter
namecraftwork
description9 engineering disciplines for AI agents — TDD, debugging, code review, planning, and multi-agent coordination. Build software like it matters.
version1.0.0
metadata[object Object]

Craft — Engineering Discipline for AI Agents

9 skills that turn AI agents into disciplined software engineers. Battle-tested on a 4-agent team shipping production software.

Skills

SkillWhat It Enforces
craftwork:test-driven-developmentNo production code without a failing test first
craftwork:systematic-debuggingNo fixes without root cause investigation
craftwork:verification-before-completionNo "done" claims without fresh evidence
craftwork:brainstormingNo code without exploring approaches first
craftwork:writing-plansNo implementation without exact file paths, code, and tests
craftwork:subagent-driven-developmentOne subagent per task, structured review between each
craftwork:dispatching-parallel-agentsSafe parallel execution for independent tasks
craftwork:requesting-code-reviewStructured diff review with security and spec checklist
craftwork:receiving-code-reviewTechnical rigor over performative agreement

Install

clawhub install craftwork

Philosophy

Agents are fast but sloppy. They skip tests, guess at fixes, claim "done" without checking, and agree with every review comment. These skills add guardrails — not to slow agents down, but to make their speed count.

Every skill follows the same pattern: stop, verify, then act.