SuperClaw by Kamir Labs#

A collection of 14 OpenClaw skills that enforce disciplined, verifiable agent workflows.

What's Included#

SkillWhen to UsebrainstormingBefore creative or feature-design workdispatching-parallel-agents2+ independent tasks that can run in parallelexecuting-plansExecute a written plan with checkpointed reviewsfinishing-a-development-branchAfter implementation, choose merge/PR/cleanup pathreceiving-code-reviewProcessing review feedback, especially ambiguous commentsrequesting-code-reviewBefore merge to verify quality and requirementssubagent-driven-developmentExecute plans via independent subagent taskssystematic-debuggingBugs and failures — root cause before fixestest-driven-developmentBefore writing implementation codeusing-git-worktreesFeature work that needs workspace isolationusing-superpowersSession start — identify the right skill firstverification-before-completionBefore any "done" claim — prove it workswriting-plansMulti-step tasks that need concrete execution planswriting-skillsCreating, editing, or validating skills

Install#

bash

npx clawhub@latest install superclaw

Philosophy#

Stop guessing. Stop claiming "done" without evidence. Stop debugging by random changes. These skills enforce the patterns that separate reliable agent work from chaos.