GITHUB AGENT SKILLS / 技能收录
给你的 AI 助手
装上专业技能
我们从 GitHub 持续收录、审核并分类优质的 Agent Skills,按职业与使用场景整理,复制一行命令即可安装到 Claude Code、Cursor 等智能体。
按职业浏览 Agent Skills
为不同岗位挑选的高频技能
按使用场景浏览 Agent Skills
从你要完成的任务出发
按创作者浏览 Agent Skills
按 GitHub Stars 排序的优秀创作者
Microsoft
@microsoftopenclaw
@openclawAnthropic
@anthropicsOpenAI
@openaiThe Apache Software Foundation
@apacheGitHub
@githubAutoGPT
@Significant-GravitasGarry Tan
@garrytan全部 Skills 482556
pk-team-create
Command: pk-team-create
pk-status
Use the standup-context skill to generate a status update: current
pk-standup
Generate today's standup: what was done, what is in progress, what comes
pk-skill-new
Use the skill-builder skill to start the interactive skill-creation workflow for $ARGUMENTS.
pk-skill-audit
Use the skill-reviewer skill to run a full 11-category audit of the $ARGUMENTS skill.
pk-route
Use the model-recommender skill, Workflow B (Task Router), to route the following task to the optimal model + version + effort: $ARGUMENTS.
pk-review
Review the current branch's changes against the project's code style and
pk-retro
Use the retrospective skill to generate a post-release blameless
pk-resume
Use the status-briefing skill to run `pk-reconcile session-start`, then
pk-release-audit
Use the release-audit skill to run the pre-release validation sweep and print
pk-owner-bootstrap
Use the owner-profiling skill to run the initial profiling interview for $ARGUMENTS.
pk-observe
Use the owner-profiling skill to record a behavioural observation: $ARGUMENTS.
pk-note
Use the note-management skill to capture a new fleeting note titled
pk-note-review
Use the note-management skill to review all fleeting notes and decide what to promote or discard.
pk-note-promote
Use the note-management skill to promote note $ARGUMENTS to a permanent artifact.
pk-model-setup
Use the model-recommender skill, Workflow D (Setup Questionnaire), to configure model access and project-level preferences.
pk-model-refresh
Use the model-recommender skill, Workflow C (Roster Refresh), to research and update the model roster from live benchmarks.
pk-lint
Read the `<!-- pk-commands BEGIN -->` ... `<!-- pk-commands END -->` block