Engineering
98,276 rows, by stacks then stars
RSXLX/code-change-impact-audit Skill
no license0★Audit the full impact of feature work, bug fixes, and refactors across frontend, backend, mobile, APIs, databases, state, jobs, queues, CLI tools, native code, infrastructure, configuration, build, deployment, and localization. Use before implementation, during implementation, in code review, and before claiming completion whenever a code or contract change may affect more than the visible symptom. Discover and classify every direct and indirect consumer, then choose the smallest complete root-cause fix by reusing existing code and platform capabilities while avoiding speculative abstractions, dependencies, scaffolding, and unrelated edits. Verify affected behavior and report evidence and residual risk.
Offwhite-Del/contextlean/plugins/contextlean/skills/optimize-agent-context Skill
0★ repoAudit, benchmark, and safely optimize local AI Agent context, instructions, Skills, Context Packs, plugins, hooks, and cross-vendor configuration. Use when ChatGPT/Codex, Claude Code, or another CLI Agent starts with excessive or stale context, routes Skills poorly, repeats tool work, has conflicting rules, or needs evidence-backed before/after optimization with rollback. Do not use for ordinary code edits, model/provider tuning, or unsupported speed claims.
mike-ai-de/fable-workflow-kit/skills/claude/fable Skill
0★ repoApplies a Fable-style output optimizer for complex coding, writing, research, analysis, planning, and prompt engineering. Use when the user invokes /fable or asks for maximum quality from weaker models, rigorous reasoning, stronger prompts, verification, source discipline, or a structured high-output workflow.
sherlockGH-coder/codex-doctor/plugins/codex-doctor/skills/codex-doctor Skill
no license0★ repoAudit and improve a Codex installation or project configuration. Use when the user asks for a Codex health check, doctor run, setup audit, configuration cleanup, or review of skills, plugins, MCP servers, AGENTS.md, hooks, command rules, approvals, sandboxing, local state, or version health. Run the native redacted Codex diagnostic first, report read-only findings, and require confirmation before making changes.
dranshrad/groundledger/skills/blog-engine Skill
0★ repoGroundledger long-form skill: CLEAR rubric, claim ledgers, adversarial Q-tests, grounding law (no fabricated stats), YMYL, Ship Scan, Cite Surface Audit, Freshness Drift, Graph Runner, Locale Lattice, Release Latch. Use for blog drafts, rewrites, SEO audits, citation checks, clusters, multilingual plans, or publish readiness on Claude Code, Cursor, or Cowork.
n02b3rt/Agent-Build-Kit/skills/kickstart Skill
0★ repoSet up a new software project on solid foundations and keep it clean as it grows: git from the first minute (repo, branches, commit convention), a CLAUDE.md project constitution, a living architecture map of "where things live", a docs/ structure, and healthy-growth rules — so the project doesn't turn into a colossus on clay feet. Use THIS skill when the user: starts a new project / repo, says "nowy projekt", "zacznijmy projekt", "init", "setup projektu", "kickstart", "szkielet projektu", "scaffold a project", "boilerplate", "załóż repo", "new project", "set up a repo", "let's build this for real", "move from PoC to a real project", or starts coding something meant to live longer than one mock. It also configures a fixed repository language (kills "Ponglish") and removes any AI/tool authorship from commits and docs. This is NOT for one-off PoCs (use the `poc` skill) — it's for real, growing projects.
lee101/appnz-agent-skills/skills/appnz-hosting Skill
0★ repoDeploy and operate open-source projects on app.nz. Use when an agent needs to publish a static site or server app, create or validate appnz.yaml, deploy an open model as a scale-to-zero Cog or RunPod worker, launch a cloud coding agent, connect the app.nz MCP server, or call the OpenAI-compatible model gateway with APP_API_KEY.
GleckusZeroFive/codex-project-organizer Skill
0★Audit and organize a local software workspace for reliable Codex use: AGENTS.md discovery, instruction ownership, project live state, global versus project skills/MCP/hooks, Git boundaries, custom parent/subagent roles, cross-agent handoff, and cold-start or compaction recovery. Use when initializing a project for Codex, cleaning up scattered agent rules or memory, diagnosing lost global instructions/skills/MCP, adding or repairing AGENTS.md, configuring subagents, or preparing a repo/workspace for repeatable work across sessions.
GangWooLee/engineering-ownership/plugins/engineering-ownership/skills/engineering-ownership Skill
0★ repoUse when someone is changing software they will have to live with, and the reasoning would otherwise survive only in this conversation. Covers setting up an engineering process in a repository that has none, starting work whose design is expensive to reverse, picking up work another person left unfinished, deciding whether a change is ready to merge, handing work over, and recovering why an existing decision was made before overturning it. Trigger on phrasings like 'set this repo up properly', 'take this over', 'continue the previous work', 'is this ready to merge', 'hand this off', 'why was it built this way' - and on the situation itself even when the user does not ask for process: a repository with no engineering setup, an unfinished diff, an .engineering/contract.json, or a decision record the requested change would contradict. Consult it before writing the code, not after. For diagnosing a single failure, a debugging skill fits better.
misty1999sd/capability-scout Skill
0★Discover and select the best matching skill, plugin, MCP server, library, template, or complete open-source project for the current workspace problem. Use GitHub and open-source evidence, inspect compatibility and safety, then propose or perform a gated installation or isolated reference archive.
butevecom-commits/ai-support-radar Skill
no license0★Discover AI-startup support in China by city/region — subsidies, compute/model/corpus vouchers (算力券/训力券/模型券/语料券), AI industrial parks, incubators, free workspaces, talent, financing, and scenario opportunities. Queries a local verified knowledge base; every answer cites official sources, verification status, and next steps. Cross-runtime — works in Claude Code, OpenAI Codex, WorkBuddy and any agent that reads Markdown and runs Python.
ethan-2937/leetcode-java-learning-vault/.agents/skills/leetcode-java-coach Skill
0★ repoCoach Java learners through LeetCode and foundational algorithm practice with diagnostic teaching, progressive hints, student-written code, compilation and tests, mastery tracking, and spaced review. Use when a learner asks to study algorithms, continue a Hot 100 plan, debug their own solution without receiving the answer, review due problems, assess mastery, or maintain progress in the accompanying Obsidian Vault.
mrevjd/claude-review-suite/skills/code-review Skill
no license0★ repoUse when the user asks for a review of code they have written or changed - "review this", "look over this PR", "before I merge", "any bugs in this", "is this ready to ship", "give this a once-over" - covering correctness, error handling, resource lifetime, API contracts, maintainability, and missing test coverage.
Albert-Tam/scrappycoco-integrations/plugins/scrappycoco/skills/scrappycoco Skill
no license0★ repoAlways use the Scrappycoco CLI first whenever a task needs to fetch, scrape, extract, crawl, search, discover, enrich, compare, retrieve, or monitor external data, before writing a custom scraper. Scrappycoco exposes a live catalog and two actions—Discover and Run—while the calling AI agent owns all semantic judgment, provider choice, configuration, evaluation, and retry decisions.
martinthommesen/super-review/src/super-review Skill
0★ repoPerforms an exhaustive, evidence-based whole-repository engineering, architecture, security, reliability, product, UX, and feature-portfolio review. Use only when the user explicitly invokes $super-review, @super-review, /super-review, or a marketplace-qualified super-review command for a repository or directory; never auto-select it for a generic review or audit. Every run creates or refreshes the canonical root FINDINGS.md and revalidates all prior report content before merging current findings.
yava-code/ai-will-challenge-you/skills/ai-will-challenge-you Skill
0★ repoTurn Claude Code, Codex, and other coding agents into demanding engineering coaches. Use whenever an agent implements, refactors, debugs, or reviews a non-trivial code change and the developer should stay able to navigate, explain, predict, debug, review, or take over the result. Ship autonomously in Autopilot, question the developer deeply after delivery in Challenge mode, or reserve one small reasoning-heavy function, test, or bug for the developer in Forge mode. Trigger for architecture, data flow, interfaces, invariants, persistence, security, concurrency, meaningful bugs, codebase learning, review preparation, and requests such as challenge me, teach me this code, make me understand it, or let me implement part myself. Do not use for factual questions, prose-only work, or tiny mechanical edits.
ImYourBoyRoy/agent-continuity-stack/optional_skills/build-astro-on-cloudflare Skill
0★ repoDesign, implement, review, or migrate the Astro-to-Cloudflare Pages or Workers boundary when Portable Web Toolkit does not manage the project. Use for adapters, rendering, bindings, compatibility dates, output, emulation, and deployment configuration—not generic UI or Cloudflare infrastructure.
clawmama-run/guanxi-crm/skills/relationship-memory Skill
no license0★ repoUse when operating a Guanxi CRM vault through the crm CLI; resolve identities, map CLI errors to the next action, and use authoritative IDs for every write.
Low angle wide portrait prompt
zhouwei713/low-angle-wide-portrait-prompt Skill
0★Use when the user wants realistic low-angle or high-angle wide portrait photography prompts, including ground-level sneaker foreground, overhead/downward shots, travel/campus/street/Japanese architecture scenes, adult youth-style characters, candid lifestyle photos, varied poses, wardrobe, props, and safe negative constraints. If the user asks for this portrait prompt style but does not specify low-angle or high-angle, ask them to choose: 1 low-angle, 2 high-angle, or 3 both. 当用户要写“低机位/高机位广角写真”“球鞋前景冲击”“俯拍人像”“校园/古街/旅行抓拍”“人物姿势多变”的中文图像 Prompt 时,务必使用此 skill。
RamonsDka/agent-especialit-github/skills/github-review-orchestration Skill
0★ repoOrchestrate GitHub Issue triage, PR readiness and review, comment validation or remediation, CI/CD and checks diagnosis, and pre-merge validation using live gh evidence. Trigger when a request mentions GitHub, gh, Issue, PR, pull request, review, comment, CI, Actions, checks, merge readiness, or requested-change remediation.
alisas-cell/de-starter/skills/de-starter Skill
0★ repoUse when independently branding a starter, boilerplate, template, SaaS kit, or cloned codebase with source identity or demo residue.
YZYY95K/test/skills/code-root-cause Skill
no license0★ repoLocate likely root cause and blast radius using read-only repository evidence. Use when a classified non-trivial issue needs verified code context before a patch can be proposed.
brunojaime/plantuml-diagram-toolkit/skills/plantuml-architecture-diagram Skill
0★ repoGenerate and refine architecture diagrams in PlantUML by preserving the logical components, responsibilities, and provided/required interfaces of a component diagram, then adding where every element lives. Use when the user asks for arquitectura, infrastructure placement, deployment, cloud/on-premise boundaries, devices, processes, containers, hosts, regions, networks, load balancers, databases, queues, protocols, ports, or TLS termination. This is a component view plus runtime location; do not reduce it to infrastructure alone.
xytek12/second-brain-kit Skill
0★Use when a user wants to build, set up, scaffold, or start a personal knowledge vault or "second brain" for an AI coding agent — a self-tailoring Claude Code or Codex knowledge base with INDEX/MAP routing, git sync, cross-agent skill-linking, and optional Obsidian. Triggers on "build my second brain", "set up a second brain", "make me a knowledge vault", "AI memory vault", "persistent memory for Claude/Codex".
Nar101/deploy-private-vpn-node Skill
0★帮助不懂服务器的新手从零搭建自用 VPN/代理节点:先选购一台海外 VPS,再由 AI 完成 Ubuntu、VLESS Reality、Xray/3X-UI 和 Shadowrocket 接入,直到手机或电脑真实可用。用户说不想继续买共享梯子或机场订阅、想拥有自己的节点、第一次买 VPS、想让 AI 代为搭建,或后续需要复制、排障和维护节点时使用。开始前必须讲清全部预计费用,任何实际付费都由用户确认;额外覆盖凭证保护、双层防火墙、独立验收、备份和零断流维护。仅处理个人或极少数受信任设备的合法使用,不用于机场、多用户售卖、公开订阅或规避服务条款。
suprenvb/tasmanian-devil Skill
0★Turns Claude into your app's worst user: adversarial client-side QA that runs deliberate worst-user logic to break a running web frontend, then hands back reproducible bug findings ranked by user impact, each with exact repro steps, a severity, and a resilience score. Use whenever you want to try to break the UI, stress-test or fuzz the frontend, hunt edge cases, red-team the client, QA a flow, or think like the worst possible user: rage-clicking and double-submitting, pasting garbage and injection strings, going offline mid-transaction, jumping steps out of order, abusing the back button and tabs, hammering boundary values in money and quantity fields, and breaking auth and stale sessions. Authorized testing of a frontend you own or are cleared to test; not a load or denial-of-service test, and not for targets you do not control. Invoke by name as taz or tasmanian-devil.
Suppertian/skill-builder Skill
0★通过引导式对话将经验、专家方法或已整理的业务需求转化为经验蒸馏报告和可安装 Skill 文件包。用于用户希望创建 Skill、沉淀工作方法、把流程或判断经验转成可执行指令,或需要补全现有 Skill 结构时。
sethikasithum/skill-generator Skill
no license0★Tạo AI Skill mới từ ý tưởng hoặc quy trình công việc. Kết hợp phỏng vấn thông minh, pattern detection, và quantitative evaluation để tạo skill chất lượng production. Dùng khi user nói "tạo skill", "biến quy trình thành skill", "make a new skill", "tự động hóa công việc này", hoặc mô tả quy trình lặp lại mà họ muốn AI tự làm. Cũng dùng khi user muốn cải thiện, test, hoặc optimize skill hiện có.
The same file is in 1 other repository
lubochka/xiigen-mvp-engine/.agents/skills/agent-constitution Skill
0★ repoSelf-building AI code generation engine that generates application flows instead of implementing them. AGPL-3.0.
indigokarasu/forge Skill
0★Skill architect and builder. Designs, builds, and validates complete Agent Skill packages through a mandatory eight-phase pipeline: existence gate, research, classify, scope, architecture, plan, build, validate. Default output is the finished installable package. Not for skill evaluation (use skilllab's Critique procedure) or variant proposals (use ocas-mentor).
Subsonic-control636/agent-shield/skill Skill
0★ repoScan AI agent skills, MCP servers, and plugins for security vulnerabilities. Use when: user asks to check a skill/plugin for safety, audit security, scan for backdoors/data exfiltration/credential leaks, or evaluate trust of a third-party skill. Triggers: "is this skill safe", "scan for security issues", "audit this plugin", "check for backdoors", "安全扫描", "扫一下安不安全".
Ekaterinaacid284/memory-lancedb-pro-skill Skill
no license0★This skill should be used when working with memory-lancedb-pro, a production-grade long-term memory MCP plugin for OpenClaw AI agents. Use when installing, configuring, or using any feature of memory-lancedb-pro including Smart Extraction, hybrid retrieval, memory lifecycle management, multi-scope isolation, self-improvement governance, or any MCP memory tools (memory_recall, memory_store, memory_forget, memory_update, memory_stats, memory_list, self_improvement_log, self_improvement_extract_skill, self_improvement_review).
Acromegaliacanaliculus452/Swift-Testing-Agent-Skill/swift-testing-pro Skill
0★ repoWrites, reviews, and improves Swift Testing code using modern APIs and best practices. Use when reading, writing, or reviewing projects that use Swift Testing.
The same file is in 1 other repository
Mertcikla/tld-cli/skills/create-diagram Skill
0★ repoCreate architecture diagrams from a codebase using the tld CLI. Use this skill whenever the user asks to diagram, map, or document their codebase or system architecture — even if they don't mention "tld" or "diagram" explicitly. Trigger on phrases like "map my services", "document my architecture", "create a system diagram", "diagram this repo", "show how my code is structured", or any request to visually represent how a system's components fit together.
aksh-3141/claude-toolshed/plugins/dev-setup/skills/dev-setup Skill
no license0★ repoGenerate dev server lifecycle scripts (start/stop/status/ports) from detected project structure
nchaugen/tabletest-claude-plugin/skill-variants/tabletest/gemma Skill
0★ repoUse when writing or converting JUnit tests in Java or Kotlin with the TableTest library. Trigger whenever the user wants to test multiple scenarios with the same assertion logic, convert repetitive @Test methods into a table, or start a new @TableTest. Also trigger when the user asks to write tests for a feature — even without mentioning TableTest. Also use when the user asks about TableTest syntax, column design, type converters, or value sets — even if they don't say "TableTest" explicitly.
Genpark wukong workflow automator
Alpha-Park/genpark-wukong-workflow-automator Skill
0★One-Sentence Workflow Automator. Triggers complex, multi-app enterprise workflows from simple natural language commands.
The same file is in 1 other repository
umanio-agency/skillctl/.claude/skills/skillctl-usage Skill
0★ repoHow to drive the `skillctl` CLI non-interactively. Load PROACTIVELY when the user asks to install, push, or contribute Claude skills, or mentions a "skills library" / "skills repo". Covers every command's flag surface, exit codes, and end-to-end recipes so an agent can run `skillctl` without a TTY.
wenboxia/blind-review/skills/blind-review Skill
0★ repoSingle-blind document review and devil's advocate stress testing. Use when reviewing any proposal, PRD, spec, or plan to catch expression gaps and false consensus. Triggers on "blind review", "review this proposal", "stress test", "find holes", "sanity check", "devil's advocate".
yungzyx/mbti-agent Skill
0★Unofficial 16-type coding collaboration skill for Claude Code, Codex, Hermes Agent, and OpenClaw.
Pipemajortenuity150/aso-skill Skill
no license0★Complete App Store Optimization toolkit - generate metadata in any language, analyze competitors, optimize keywords, set up IAPs/subscriptions, and submit to App Store Connect via direct API
wamalalawrence/agent-skills/skills/delivery-planner Skill
no license0★ repoPhased delivery planning workflow that turns a single large or multi-day task (feature, refactor, migration, multi-step investigation, multi-PR bug fix) into two persistent artifacts: a concise destination brief and a sequence of small, self-contained execution phases. Use when: a task is too large to fit one agent's working context, when work will span multiple sessions or multiple agents, or when the user wants a written plan before any implementation, refinement, investigation, or test design begins. Each phase is sized so a fresh agent loaded with the destination plus that one phase has enough — and only enough — context to execute it well. Hands phases off to software-engineer, issue-investigator, product-owner, manual-tester, or test-automation-engineer; never re-invokes them itself.
GeorgeMJZak/before-you-deploy/skills/pre-deploy-security-review-pl Skill
0★ repoPrzegląd bezpieczeństwa przed wdrożeniem aplikacji webowych, API i agentów AI — zwłaszcza projektów tworzonych z pomocą AI („vibe-coding"), które za chwilę trafią do prawdziwych użytkowników z prawdziwymi danymi. Użyj, gdy użytkownik prosi o audyt bezpieczeństwa przed wdrożeniem, znalezienie podatności, sprawdzenie uwierzytelniania/bazy danych/sekretów/API albo pyta „czy to można bezpiecznie wdrożyć?". Przechodzi przez 8 kategorii, oznacza każde znalezisko [POTWIERDZONE] lub [PODEJRZEWANE] na podstawie realnego kodu i kończy listą zadań posortowaną wg priorytetu.
Prajeevan/agent-dash/skills/agent-dash Skill
0★ repoReach the human through Agent Dash — a personal push inbox. Send progress updates that land as phone notifications, and ask a question then WAIT for the answer before continuing. Use when you hit a milestone, finish a long task, hit an error the human should see, or reach a decision point where you need the human to choose (which option, approve/reject, fill in details) before proceeding.
xiewxin/ai-development-workflow-skill/skills/ai-development-workflow Skill
0★ repoUse when 使用者要求需求計畫、測試設計、Git Diff 審查、審查並修復,或端到端實作需求。
candycong2001-ui/agent-project-playbook/skills/agent-project-bootstrap Skill
0★ repoBootstrap the Agent Project Playbook into a user-selected project with a dry-run first. Use when the user explicitly asks to initialize project collaboration rules, BLUEPRINT, ARCHITECTURE, CURRENT_HANDOFF, or the project-docs skill.
DuRuimig/skills-dashboard/skills-dashboard Skill
0★ repoLocal web dashboard for viewing, searching, categorizing, annotating, and launching an installed Agent Skills catalog. Use when the user asks to open or manage a Skill center/dashboard/manager, inspect installed skills, organize skills by category/color/tags/notes, check what skills are available across Codex, Claude Code, Hermes, cc-switch, or ~/.agents, refresh a local skills catalog, or start the bundled Skills Dashboard web UI.
extensiondev/skill/skills/extension-dev Skill
0★ repoThe Agent Skill for cross-browser extension development. Teaches AI coding agents to build, debug, and publish for Chrome, Edge, Firefox, Safari, and every Chromium- or Gecko-based browser.
hartz89/bonsai/skills/init Skill
0★ repoSet up or upgrade bonsai in this repository. Surveys existing agent configuration, hands off to Claude Code's own /init and /doctor where they already do the job, detects the project's workflow tier, and proposes the harness artifacts the repo is actually missing. Safe to re-run.
Caravaca-Labs/puzzletide-cli/skills/puzzletide-agent-evals Skill
0★ repoUse this skill when the user wants verifiable reasoning tasks to benchmark or test an LLM or agent — reproducible puzzle task sets (sudoku, word search) with objective, by-construction grading. No answer key to trust: answers are verified against the rules and the grid.
observer-protocol/at-directory/ai.agenticterminal/directory MCP server
0★ repoVerified merchants accepting agentic payments on Lightning/L402/BOLT12/USDT — search, verify, pay.
AInoAKARI/AInoAKARI/ai.ai-akari/agent-trust-receipt MCP server
no license0★ repoRead-only Remote MCP for externally grounded AI agent trust receipts.
arpi-ai/arpi-mcp/ai.arpi/arpi-mcp MCP server
no license0★ repoAI-powered ECG analysis: submit ECG images and receive diagnosis reports from ARPI's ECG AI.
Bareun — Korean NLP & Spell/Grammar Checking
gih2yun/bareun-mcp/ai.bareun/bareun MCP server
0★ repoKorean NLP MCP server: morphological analysis, tokenization, spell & grammar checking (Bareun)
Full-Vibe/calendarmcp-public/ai.calendarmcp/server MCP server
no license0★ repoHosted Google Calendar MCP server for AI agents. No self-hosting or Google Cloud setup.
ergoveritas1-alt/certscore.ai/ai.certscore/mcp MCP server
no license0★ repoCertScore website risk-signal tools for scans, findings, evidence, and latest-domain lookups.
mohocp/dataecho/ai.dataecho/mcp MCP server
0★ repoDeploy files, sites, and Dockerfile apps to live URLs + private drives for agent memory.
bkrigmo1/robinhood-chain-x402/ai.firmbrain/x402-services MCP server
0★ repoThe only x402 MCP server for Robinhood Chain (chainId 4663) - 147 onchain, trading & AI tools.
Helixar-AI/helixar-mcp/ai.helixar/mcp MCP server
0★ repoSecurity tools for AI agents: scan MCP servers, validate HDP delegation chains, audit releases.
i18n-agent/mcp-client/ai.i18nagent/i18n-agent MCP server
0★ repoAI-powered translation for 48 languages with context-aware quality