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。

01 /

按职业浏览 Agent Skills

为不同岗位挑选的高频技能

全部职业
02 /

按使用场景浏览 Agent Skills

从你要完成的任务出发

全部场景
03 /

按创作者浏览 Agent Skills

按 GitHub Stars 排序的优秀创作者

全部创作者

全部 Skills 482556

vivekvjnk
vivekvjnk/three-wise-monkeys

Boundary_Reconciliation

暂无简介

★ 0
vivekvjnk
vivekvjnk/three-wise-monkeys

Boundary_Alignment

暂无简介

★ 0
JonasThowsen
JonasThowsen/.dotfiles

ash-framework-idioms

Write idiomatic Ash framework code using intentful actions, code interfaces, relationship management, DSL-first modeling, and framework-native changes/preparations instead of ad hoc helper layers. Use when creating or reviewing Ash resources, domains, actions, queries, or Phoenix integrations.

开发工程师
★ 0
vivekvjnk
vivekvjnk/three-wise-monkeys

Assembly_Integration

暂无简介

★ 0
vivekvjnk
vivekvjnk/three-wise-monkeys

project-structure

Materialized view of observed project reality for user project. Discovers and documents project root, modules, worktrees, branches, system/module boundaries, SCUDs, component libraries, circuit sources, and evaluation logs. Trigger when navigating the project filesystem, locating artifacts, or understanding current project structure.

★ 0
iyulab
iyulab/claude-plugins

telemetry-az

Analyzes Azure Application Insights telemetry since the last run to surface defects, performance regressions, and feature-drop signals, tracked run-over-run as a quality trend (error-rate/p95/issue-count direction, with closed-issue recurrence detection), plus a report-only user-analytics reading (active-user growth, feature/page preference shifts, engagement). Triages findings against project philosophy, files issues only for threshold-crossing ones, and leads every report with a both-purpose Trend section backed by a 12-run history.

数据分析师分析与可视化
★ 0
iyulab
iyulab/claude-plugins

ship

Takes finished work out — bumps the version, commits, pushes, and watches the CI run to completion, reporting what failed if it fails. Checks first whether now is the right moment — if remaining backlog work would touch the same consumer-facing surface, it says so and asks rather than publishing a version the next few items obsolete.

★ 0
iyulab
iyulab/claude-plugins

run-cycle

Executes adaptive iterative development cycles — each cycle is a self-contained plan/execute/verify/reflect loop that reshapes the roadmap and derives emergent follow-on scope from its own output

★ 0
iyulab
iyulab/claude-plugins

resume

Opens a work session — reads the continuity docs and recent state, synthesizes where things actually stand, proposes a re-prioritized next scope with any derived prerequisites or objections, and settles the decisions `/iyu:handoff` flagged the way a capable employee would — grounded options, trade-offs, and a recommendation that becomes the provisional decision, recorded at once for the owner to correct rather than answer. Only external blockers, owner-gated acts, and criterion gaps are left waiting on the owner. Then stops — execution is a separate, explicit step.

人力资源
★ 0
iyulab
iyulab/claude-plugins

mindset

Applies a maintainer's critical-but-constructive judgment — weighs a proposed feature, design change, or instruction against the project's own philosophy, architecture, and design docs, names any conflict instead of silently complying or refusing, and flags scope creep and boundary leaks.

设计师视觉设计
★ 0
iyulab
iyulab/claude-plugins

issue-triage

Supplies decision matrices and reply templates for accepting, adapting, deferring, redirecting, or declining an external issue, feature request, or pull request. Use when deciding whether to accept or merge input from outside contributors, or how to respond to its author.

★ 0
iyulab
iyulab/claude-plugins

handoff

Checkpoints a work session — at its close (the main use), mid-session before clearing context or switching focus, or at the end of a `/iyu:run-cycle` run. Records where things stand, migrates completed work out of the continuity docs into the history index, derives and stages the next scope, flags (not briefs) the decisions waiting on a human for `/iyu:resume` to brief when the next session opens, and commits everything it can confidently attribute (this session's work, pre-existing uncommitted leftovers, and its own doc edits) toward a clean `git status` — never pushing, and reporting rather than guessing at anything ambiguous. Produces the document a fresh session (or another person) resumes from, so nothing depends on the current conversation surviving.

★ 0
iyulab
iyulab/claude-plugins

backlog-discover

Discovers and judges new backlog phases for a project's ROADMAP.md like an owner-manager — runs a convergent/emergent research playbook (vision-gap, trend/competitive research, domain-practice and academic scans, stewardship inspection of dependencies/config/docs/repo hygiene, telemetry reuse, active dogfooding, plus emergent sessions rotated in by symptom) on a persistent per-activity cadence, then synthesizes the findings into a state-of-the-product diagnosis, an evidence-grounded importance ranking, and a dependency-ordered now/next/later staging toward the vision. Always produces a proposal document; adopts into ROADMAP.md only the items the project's own evidence grounds and that stay inside its declared role, and leaves everything else in the proposal for a human.

产品经理资料检索与研究
★ 0
ContextHelp
ContextHelp/ContextHelp

tokenizer-step

Mock step

★ 0
ContextHelp
ContextHelp/ContextHelp

my-extractor

Mock step

★ 0
ContextHelp
ContextHelp/ContextHelp

ctxt-knowledge-retrieval

Train and prompt AI agents to use ctxt for deterministic knowledge retrieval with semantic fallback, provenance tracking, and reusable memory capture. Use when building agent loops over ctxt, debugging weak retrieval quality, or sharing insights, trial-and-error outcomes, and decisions across agents and sessions.

资料检索与研究Skill与智能体开发
★ 0
ContextHelp
ContextHelp/ContextHelp

pii-mask

Masks email addresses and phone numbers in captured text

邮件与沟通
★ 0
ContextHelp
ContextHelp/ContextHelp

ctxt

Work with the ctxt CLI — an agentic context brain for capture, search, and composition on top of dPKMS (decentralized personal knowledge substrate). Automatically applies when: task involves capturing content, searching knowledge, composing briefs/plans, or managing focus profiles via the ctxt CLI. Use when: capturing content to knowledge base, running semantic search, generating compositions, managing config/profiles, or operating the dPKMS background worker. Description checklist: - [x] What it does: context-as-a-service CLI for humans and AI agents - [x] When to auto-activate: any ctxt/dpkms CLI operation - [x] Key surfaces: capture, search, compose, config, registry, entity - [x] Local-first, offline-capable, scriptable

Skill与智能体开发
★ 0