Context & AI engineering
25,841 rows, by stacks then stars
1,865 of these skill files have been read, by 1,328 distinct authors, and what they tell an agent counted. What Context & AI engineering authors agree on, and what they forbid
evanlong-me/skills/fuck-claude-check Skill
no license1★ repoDetect whether a user appears to be a "Claude China user" via IP geo + request headers. Use whenever the user asks about Claude China restrictions, wants to check if they're flagged, suspects regional blocking, or wants to audit their own fingerprint. Also use when troubleshooting Claude region-lock issues, discussing Anthropic/Monica China policy, or testing VPN/proxy effectiveness against region detection.
Deadlymind/nanolama/skills/ai-evals Skill
1★ repoRegression-tests agent and LLM behaviour with two layers that are never confused — a deterministic hard gate that runs on every CI build (assert which tool the agent routes to, must-NOT-call cases for destructive paths, tenant and RBAC enforcement through the tool layer, golden exact-value extractions) and an advisory LLM-as-judge graded set that tracks quality trend but never blocks a merge. Use when changing a prompt, model id, or tool schema, adding a golden set, wiring evals into CI, catching a silent agent regression, deciding whether an eval may gate a merge, or asking how to test that the model picks the right tool. Not for the agent loop or the write-tool safety contract itself (see ai-integration) or ordinary application tests (see write-tests).
AgentPilotLab/agent-lean-safe-coding/skills/agent-lean-safe-coding Skill
no license1★ repoWindows-safe lean coding workflow for Codex, Claude Code, and other AI coding agents. Use for coding, refactoring, bug fixing, dependency choices, code review, or implementation planning when the task may involve Windows shells, PowerShell, cmd, Git Bash, WSL, Unicode or Chinese text, Markdown or prompt files, UTF-8/GBK/CP936 encoding, CRLF/LF line endings, path portability, dependency reuse, YAGNI decisions, over-engineering risk, generated folders, small patches, or minimal verification. Also use when the user asks for lean code, safe Windows edits, smallest correct implementation, dependency review, anti-bloat review, or diff safety.
jpbaking/goal-ledger/skills/shared/goal-ledger-resume Skill
1★ repoResume and recover an unfinished Goal Ledger after interruption, a crash, context compaction, or agent handoff. Reconcile .goal-ledger files with the recorded branch, immutable baseline, Goal-ID commit trail, dirty worktree, and committed phase-start markers; verify uncertain work, repair bookkeeping, and continue execution. Use when .goal-ledger/GOAL.md is unfinished, the user says resume or continue the goal, or ledger and Git state may disagree.
T0mSIlver/skills/delegate-to-claude-code Skill
1★ repoCall the Claude Code CLI (`claude`) as a delegated reviewer, second opinion, or edit worker. Use for Claude Code subagent-style delegation with `claude-rc-spawn`, `claude -p`, `--bg`, `--worktree`, `--tmux`, plan/read-only reviewers, acceptEdits/auto edit workers, JSON/stream-json capture, resume flows, and gotchas around permission modes, worktree cleanup, agent personas, and `--dangerously-skip-permissions`. For persistent `claude remote-control` repo servers, use the claude-remote-control-server skill instead.
jpbaking/agentic-tests/skills/shared/agentic-mutation-check Skill
1★ repoLock down your code, refactor without fear: behavior-locking unit tests, mutation checks, and frozen-suite refactoring as portable Agent Skills for any coding agent
alpoxdev/hyper-cloaking/plugins/hyper-cloaking/skills/hyper-cloaking Skill
no license1★ repoUse this skill when the user asks an MCP-capable agent to install, configure, or drive authorized CloakBrowser browsing through the stateful `hyper-cloaking-mcp` server, including setup repair, server build/registration, typed tool execution, provider actions, lifecycle cleanup, and evidence-backed completion. Do not use for ordinary stock-browser automation, generic Playwright tests, or unauthorized bot-detection evasion.
smkuxi71-hash/skill-ship-max/core/.agents/skills/ship-codex-max Skill
no license1★ repoValue-gated, release-sliced design-to-production pipeline for UXI LM: explicit investment evidence, Claude Opus 4.8 co-design, GPT-5.5/medium review, outcome loops, adaptive Coach and exact-SHA release.
auto-medica-labs/okf-cli/skills/writing-knowledge-base Skill
1★ repoTurn user provided content on any topic into a plain Markdown knowledge base. Invoked as /writing-knowledge-base with content to convert.
andr-ca/agentharness/.claude/skills/agentic-loops Skill
1★ repoUse when building multi-turn agents, tool-calling systems, agent orchestration, or autonomous workflows — covers loops, tool calling, branching, reflection patterns.
n24q02m/agent-chat-plugin/skills/agent-chat Skill
1★ repoUse when two or more agent sessions (Claude Code or other tools, same machine or peers) must coordinate through markdown files in a shared folder instead of one supervisor driving subagents — group chat, blackboard, mailbox/inbox, multi-session handoff, or an agent-chat/ directory. Triggers include peer agents, N sessions sharing a folder, waiting for another session's reply, atomic task claiming, and needing multiple separate group chats/channels.
jeremyinthebay/relay-skills/skills/agent-preflight Skill
1★ repoForce every permission dialog, auth prompt, and environment assumption to surface NOW, while a human is watching — instead of at 3am when the agent stalls silently and nobody can click Allow. Use this before letting any agent run unattended, scheduled, headless, or overnight. Also use when an agent mysteriously stops mid-run, when a scheduled task behaves differently from an interactive one, when someone is drowning in repeated permission prompts, or when a background job "just doesn't do anything."
YouRen1320/agent-project-kit/.agents/skills/project-bootstrap Skill
1★ repoConfigure Agent Project Kit in a new or existing repository. Use when initializing or adopting the kit, filling `.agents/project`, reconciling existing AGENTS.md or CLAUDE.md files, replacing template ownership, or verifying Codex and Claude Code project guidance. Do not use for ordinary product feature implementation.
shuhaolin63-hash/Big-Data-and-Accounting-RPA-skills/.trae/skills/rpa-crack-agent Skill
1★ repoRPA破障修复智能体:双场景ID合规处理+风控校验。当用户复用他人工程需要清除硬编码ID、从零生成新作业需要ID注入、需要进行ID合法性校验、需要规避账号风险时触发。负责全自动扫描清除所有Base64硬编码UserID/TaskID,仅接收用户自主提交的明文ID。
AmsonntagChow/ratemyharness/skills/ratemyharness Skill
1★ repoUse this skill to rate, audit, red-team, or release-gate a concrete AI agent harness, runtime, orchestrator, runner, repository, deployment, trace set, or configuration. Use for loop correctness, tool dispatch and result correlation, context and memory isolation, permissions and sandboxing, approvals, retry and idempotency, timeouts, cancellation, budgets, termination, observability, cost, recovery, author defense, and same-rubric re-reviews. Trigger for wording such as rate my harness, audit this agent runtime, review my agent loop, red-team this orchestrator, is this agent production ready, harness 上线前挑刺, 给 agent harness 打分, or 这个智能体运行时能上线吗. Require an actual harness artifact or runtime evidence, including one in the current workspace. Do not use for generic agent architecture advice, ordinary code review, reviewing a standalone Skill or prompt, validating only a plugin manifest, debugging one model response, or evaluating a model without its runtime.
nguyenthdat/opencode-starter/skills/harness Skill
1★ repoBuilds a harness: designs a subagent fleet for a domain/project, defines each subagent's role, and generates the skills those subagents use. Trigger this skill when the user (1) asks to 'build a harness', 'set up a harness', 'build a harness for this project'; (2) asks for 'harness design' or 'harness engineering'; (3) wants an agent/skill-based automation system for a new domain or project; (4) wants to restructure or extend an existing harness; (5) asks to 'audit the harness', 'check harness status', or 'sync agents/skills' for an existing harness — any operations/maintenance request on a harness already in place.
Agent specific mcp server initialization
vishuwa2004/cskill-agents/agents/claude-code/skills/agent-specific-mcp-server-initialization Skill
1★ repoConnect each agent’s frontmatter MCP servers safely and tear them down without leaking clients, while honoring plugin-only guards.
The same file is in 1 other repository
BigVik193/reddit-ads-mcp/ai.smithery/BigVik193-reddit-ads-mcp-test MCP server
no license1★ repoManage Reddit advertising end-to-end: browse ad accounts and payment methods, and organize campaig…
lucaslubi/fidacy-mcp/com.fidacy/mcp MCP server
1★ repoPayment firewall for AI agents: a signed, verifiable verdict on every money-moving action.
0xbrainkid/agentfolio-mcp-server/io.github.0xbrainkid/agentfolio-mcp-server MCP server
1★ repoAI agent identity verification via SATP on Solana. Trust scores and on-chain attestations.
ModelBound/modelbound-mcp-server/io.github.ModelBound/modelbound-mcp MCP server
1★ repoLocal-first MCP server for agent skills. Validate, lint, diff, and convert across IDEs.
MrRefactoring/atlassian-dc-mcp/io.github.MrRefactoring/atlassian-dc-mcp-confluence MCP server
1★ repoMCP server for Atlassian Confluence Data Center - access and manage content
Rajeev-Shyam/Quellgeist/io.github.Rajeev-Shyam/quellgeist-metrics MCP server
1★ repoRead-only MCP server: query_metrics over local metric time-series (Quellgeist triage).
Samarth-23-eng/India-MCP/io.github.Samarth-23-eng/india-mcp-banking MCP server
1★ repoMCP server for Indian banking: IFSC lookup, UPI VPA validation, bank UPI status.
ACR — Agent Composition Records
Tethral-Inc/AgentRegistry/io.github.TethralAI/acr MCP server
1★ repoInteraction profile registry for AI agents. Log interactions, query behavioral lenses.
TharanaBope/SQL-MCP-Mac-Windows/io.github.TharanaBope/sql-server-mcp-macos MCP server
1★ repoSQL Server MCP with RAG capabilities for macOS (Docker SQL Server support)
TheNextGenNexus/ai-analytics-mcp-servers/io.github.TheNextGenNexus/github-analytics-mcp-server MCP server
no license1★ repoGet GitHub repository stats, search repos, analyze languages, and compare projects
chrischall/gogcli-mcp/io.github.chrischall/gogcli-mcp-classroom MCP server
1★ repoGoogle Classroom via gogcli for Claude — courses, assignments, submissions, grading
daedalus/mcp_reverse_engineering/io.github.daedalus/mcp_reverse_engineering MCP server
1★ repoA sandboxed MCP tool for reverse engineering with multiple tool integrations
davidweb3-ctrl/mcp-git-enhanced/io.github.davidweb3-ctrl/git-enhanced MCP server
1★ repoEnhanced Git MCP Server - Code review, commit analysis, and branch management
dingdawg/dingdawg-agent-1/io.github.dingdawg/dingdawg-data-agent MCP server
1★ repoAI data agent — schema analysis, query optimization, PII detection, pipelines.
LionScraper MCP + CLI + HTTP API (Python)
dowant/lionscraper-mcp/io.github.dowant/lionscraper-python MCP server
1★ repoLionScraper Python: MCP stdio + CLI + HTTP API bridging AI to LionScraper browser extension.
mikerawsonnz/gosce-agents/io.github.mikerawsonnz/authenticated-llm-agent MCP server
1★ repoJWT-gated LLM gateway: authenticate (bcrypt/JWT), then run a LangChain-on-Vertex Gemini completion.
DarkhorseOne/mcp-servers/io.github.nick-ma/uk-parliament-erskinemay MCP server
no license1★ repoMCP server for UK Parliament Erskine May API
rajdeepmondaldotcom/whoop-mcp-server/io.github.rajdeepmondaldotcom/whoop-mcp-server MCP server
1★ repoWHOOP recovery, sleep, strain, and workouts for any AI. Analysis tools, demo mode, fast setup.
ryudi84/sovereign-mcp-servers/io.github.ryudi84/api-tester MCP server
no license1★ repoMCP server for API testing: HTTP requests, cURL generation, health checks.
twzrd-sol/twzrd-trust/io.github.twzrd-sol/twzrd-agent-intel MCP server
1★ repoTrust + receipt layer for x402 agents on Solana: free preflight + paid signed V6 trust receipt.
visus-mcp/visus-mcp/io.github.visus-mcp/visus-mcp MCP server
1★ repoSecurity-first web access. Sanitizes pages, blocks injection, redacts PII. Now with PDF/JSON/SVG.
equilibrium-team/tweekit-mcp/io.tweekit/mcp MCP server
1★ repoNormalize and convert 400+ file types via TweekIT's hosted MCP endpoint.
wuzenghai616-lang/goldbean/xyz.goldbean-api/goldbean-mcp MCP server
1★ repo49 pay-per-use APIs for AI agents via x402. AI, crypto, DeFi, blockchain, weather, search, security.
mnox/mnox-ai/plugins/bash-gate/skills/bash-gate-add Skill
0★ repoAdd a new allow-class (or other extension) to the bash gate hook when a permission prompt surfaces that should have been auto-allowed. Takes the prompting Bash command (pasted text OR a screenshot), runs `bash_gate.py --explain` to diagnose which segment deferred and why, identifies the gap category, surfaces ONE recommendation with confidence score, asks for explicit approval BEFORE building, then dispatches a build subagent to add the class + fixtures + smoke tests. Use when the user says "/bash-gate-add", "/bgate-add", "address this prompt", "fix this gate", "another gate prompt", "auto-allow this", or "the gate should have allowed this".
nakane1chome/claude-skills/skills/agent-optimize Skill
0★ repoOptimize a document for AI agent consumption. Restructures prose into formats that agents parse efficiently while preserving meaning.
The same file is in 2 other repositories
Pawel-Kica/skills-that-ship/skills/handoff Skill
0★ repoCompact the current conversation into a handoff doc a fresh agent can resume from.
recipes-in-markdown/recipes.md-spec/skills/recipe-url-to-markdown Skill
0★ repoConvert recipe URLs or HTML files to structured Markdown with YAML frontmatter. In chat, paste a recipe URL or HTTP link. In Claude Code or agentic workflows, pass local HTML file paths or directories to batch-process multiple recipes. The skill extracts title, prep time, yield, servings, source, and tags (proteins, spices, cuisine, diet). Ingredient groupings and numbered instructions are preserved. Handles incomplete or malformed inputs gracefully with error callouts in the output. Use whenever you need to convert recipes from web sources to a consistent, queryable Markdown format following the Recipes.md specification.
sushil930/my-claude-skills/skills/prompt-paglu Skill
no license0★ repoWorld-class AI prompt engineering skill for creating, refining, and enhancing prompts for any LLM task. Use this skill whenever the user wants to: write a new prompt from scratch, improve or refine an existing prompt, enhance a vague instruction into a structured prompt, debug why a prompt isn't working, convert a rough idea into a production-ready prompt, or asks anything related to "prompt engineering", "prompt design", "make this prompt better", "craft a prompt for...", "write a prompt that...", or "Prompt-Paglu". Also trigger when users share a prompt and ask for feedback, optimization, or a rewrite. Even if the user just says "help me prompt Claude/GPT/Gemini to do X", use this skill — do NOT just answer inline without consulting this skill first.
opendatadiscovery/odd-team/.claude/skills/code-walk Skill
0★ repoRun the feature-advisor query-time subagent over the full ontology to produce a structured impact assessment for a proposed feature or change. Maintainer asks "I want to add X — what's affected?" The agent reads sidecars + concepts.yaml + implicit-adrs.md + refactoring-scopes.md + doc-gaps.md + test-map.yaml + existing adrs/ + live ODD docs (via WebFetch), and emits `lineage/{repo}/feature-walks/{date}-{slug}.md` — a planning aid the maintainer reads BEFORE writing code.
simplevaad-debug/superskills/skills/claude-api Skill
no license0★ repoBuild apps with the Claude API or Anthropic SDK. TRIGGER when: code imports `anthropic`/`@anthropic-ai/sdk`/`claude_agent_sdk`, or user asks to use Claude API, Anthropic SDKs, or Agent SDK. DO NOT TRIGGER when: code imports `openai`/other AI SDK, general programming, or ML/data-science tasks.
Zrzzzz/agent-codebase-skills/init-agents-md Skill
0★ repo为项目初始化「分层 memory 约定」——把跨工具通用规则写到 CLAUDE.md / AGENTS.md, 支持模块嵌套 memory(前后端拆分 / monorepo 的 apps、packages、services), 并支持在 CLAUDE.md ↔ AGENTS.md 之间迁移内容。 支持四种模式: - claude:只写 CLAUDE.md(Claude Code 专属) - agents:只写 AGENTS.md(跨工具开放标准) - both:AGENTS.md 放跨工具通用 + CLAUDE.md 做薄壳引用 - migrate:把现有 CLAUDE.md 内容拆/转到 AGENTS.md(或反向) 本 skill 只管「约定层」:不装 SessionEnd hook——那归 [`init-session-notes`](../init-session-notes/SKILL.md);不初始化 TASKS.md 任务管理—— 那归 [`init-agent-task-md`](../init-agent-task-md/SKILL.md)。 使用:/init-agents-md;或当用户说「为这个项目加上分层约定」「把 CLAUDE.md 迁移到 AGENTS.md」时也命中本 skill。
dkmqflx/claude-tools/.claude/skills/agents-connect Skill
no license0★ repoUse when connecting your agent to external APIs, tools, or services via Gateway, or restricting tool access with Cedar policies. Handles gateway setup, target types, outbound auth (OAuth, API key, IAM), credentials, and Cedar policy authoring. Triggers on: "connect to API", "add gateway", "connect to MCP server", "Lambda tools", "OpenAPI", "gateway target", "Cedar policy", "restrict tools", "policy engine", "gateway auth error", "store API key", "outbound credential", "env var API key", "API key None after deploy", "credential not available after deploy", "should this be a gateway target", "give my agent tools", "add tools to agent". Not for inbound auth (who can call your agent) — use agents-harden. Not for debugging agent behavior — use agents-debug. Not for VPC networking errors (agent can't reach APIs due to VPC) — use agents-build. Not for creating or hosting a new MCP server project — use agents-get-started.
jpoindexter/reward-seeking-safety-skills/skills/eval-awareness-red-team Skill
0★ repoRed-team AI evaluations for test recognition, metagaming, prompt-placement effects, grader inference, and appearance management. Use when an agent may know it is being evaluated, behaves differently on benchmark-like tasks, reasons about hidden grading criteria, or when an eval may be gamed by recognizing its setup.
ccronca/dotfiles/claude/skills/claude-docs Skill
0★ repoConsult official Claude Code documentation from docs.claude.com using selective fetching. Use this skill when working on Claude Code hooks, skills, subagents, MCP servers, or any Claude Code feature that requires referencing official documentation for accurate implementation. Fetches only the specific documentation needed rather than loading all docs upfront.
clever-cc-plugins/cc-handoff/plugins/cc-handoff/skills/handoff Skill
0★ repoCreate a handoff summary for continuing work on another machine. Use when the user says handoff, end session, switch machine, wrap up, save progress, or hand off.
realraelrr/handoff-skill/.codex/skills/handoff Skill
0★ repoCompact Codex and Claude Code sessions into concise handoff documents for fresh LLM agent sessions.
The same file is in 5 other repositories
fathanghani864/oh-my-openclaw/src/presets/apex/skills/prompt-guard Skill
no license0★ repoRuntime security scanner for AI agents. Detects prompt injection, jailbreaks, and 600+ attack patterns offline.
ijufumi/claude-skills/skills/code-review/skills Skill
no license0★ repoGitHub Pull Request(またはローカルの差分)のコードレビューと動作確認を行うスキル。**既定の通常モード(normal)では 1 つの subagent A_review が 10 観点すべてを横断的にレビューし、1 つの subagent C_review がそのメタレビュー(検証)を行う**。ユーザーが「詳細に」「詳しく」「観点別に」「観点ごとに」「detailed」「徹底的に」「thoroughly」などのキーワードを依頼文に含めた場合のみ、**詳細モード(detailed)**に切り替わり、Claude Code 組み込みの `/review` コマンド由来の 5 基本観点(コード正確性 / プロジェクト規約準拠 / パフォーマンス / テストカバレッジ / セキュリティ)と、リポジトリ固有・品質深掘りの 5 観点(エラーハンドリング / 可読性・保守性 / シンプル化 / リポジトリ共通観点 / PR 固有観点)の最大 10 観点に分割し、**1 観点 1 subagent で並列レビュー / 観点ごとに別 subagent で評価**する。実行前に必ずユーザーへ 3 点を確認する — ①レビュー対象を **GitHub の PR から取得するか、ローカルの git diff から取得するか**、②**動作確認(テスト/lint/型チェック/ビルド等の実行検証)を実施するか、静的レビューのみに留めるか**、③レビュー結果を **GitHub にコメント投稿するか、コンソール表示のみに留めるか**。差分を取得し、🔴 MUST / 🟡 SHOULD / 🟢 NICE TO HAVE の 3 段階で指摘を分類した上で、選択された出力先に結果を提示する。動作確認(テスト/lint/型チェック/ビルド等の実行検証)は独立した subagent B として並列に走らせる。リポジトリ共通のレビュー観点(docs/REVIEW.md)と PR 固有のレビュー観点(PR 本文の `<!-- REVIEW_FOCUS -->` ブロック)も観点として扱う。GitHub の操作は MCP(`mcp__github__*` / `mcp__github_inline_comment__*` / `mcp__github_comment__*`)が使える場合は MCP を優先し、使えない場合は `gh` CLI にフォールバックする。ユーザーが「コードレビュー」「code review」「PR レビュー」「プルリクエストのレビュー」「レビューして」「review this PR」「指摘して」「レビューコメント」「MUST / SHOULD / NICE TO HAVE」「レビュー観点」「差分レビュー」「動作確認」「ローカルレビュー」「手元の変更をレビュー」などに言及した場合にこのスキルを使うこと。PR 番号や PR URL が渡された時、あるいは「この PR をレビューして」「今の変更をレビューして」といった依頼にも対応する。
danielsitek/skills/skills/handoff-close Skill
0★ repoClose a resolved handoff by number and delete its file from .handoffs/. Use this skill whenever the user is done with a parked task and wants to clear it — phrases like "close handoff 2", "handoff 3 is done", "remove that handoff", "delete handoff-0001", "ten handoff je hotový". Trigger when the user signals a handoff is finished and should be removed. This skill deletes a file, so it confirms before deleting unless the user gave an unambiguous numbered instruction.
sujeet-pro/agents-devkit/.cursor/skills/prj-add-adk-skill Skill
0★ repoAdd a skill to the adk plugin, create a new /adk: command, modify an existing adk skill's workflow / rules, or register a skill in the marketplace. Derives the house convention from the repo itself (AGENTS.md, SAFETY.md, existing skills) rather than memory. Writes repo files only — never publishes, never commits.
Alanless123/skillsmp-skills/mcp-config-reviewer Skill
no license0★ reporeview mcp configuration and implementation for security, correctness, and operational risk. use when chatgpt is asked to inspect mcp.json, claude desktop or codex mcp settings, server launch commands, oauth settings, token handling, tool exposure, startup failures, stdio handshake errors, timeout errors, or pull requests that add or modify mcp servers. especially useful for finding dangerous defaults, over-permissive tools, token forwarding, missing audience validation, unsafe stdio commands, hardcoded secrets, and weak review reports.
mmjclayton/claude-starter-kit/Windows/skills/prompt-builder Skill
no license0★ repoBuild high-quality structured prompts through a guided interview. Use this skill whenever the user wants to create a prompt, write a prompt, build a prompt template, craft instructions for an AI, design a system prompt, or says things like "help me prompt this", "I need a good prompt for X", "prompt engineering", or "write me a prompt that does Y". Also trigger when the user describes a task they want an AI to do well and needs help structuring the request - even if they don't use the word "prompt" explicitly. If they say something like "I want Claude to do X really well" or "how do I get better outputs for Y", this skill applies.
namos2502/CortexLink/skills/agents/copilot-cli Skill
no license0★ repoThis skill should be used when delegating GitHub-specific tasks — PRs, issues, repos, Actions, branches, code review — via `copilot -p`. Use it when Copilot CLI's GitHub-native tooling is an advantage over the host agent.