Context & AI engineering
25,745 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
George930502/curated-codex-skills/skills/prompt-review-and-dispatch Skill
3★ repoAudit a GPT-5.6 prompt, clarify purpose through native clickable questions, polish it, require explicit approval, then continue the exact approved prompt in the current Codex conversation by default. Use when the user wants a review-before-execution prompt workflow or explicitly asks to send work to another Codex task.
madnpc/skills/skills/handoff Skill
no license3★ repo适用于将未完成任务及其上下文移交给新的代理或会话。Use when handing an unfinished task or its context to a fresh agent or a new session. 中文关键词:工作交接、会话交接、上下文交接、任务移交、续接任务。
rmonier/agent-smith/.agents/skills/skill-creator Skill
3★ repoCreates or updates vendor-neutral Agent Skills for repeated executable actions. Use when OKF, AGENTS.md, user requests, or repeated agent work reveal a reusable workflow, scriptable operation, tool integration, validation procedure, or action pattern that should live under .agents/skills/ instead of context documents.
ligyDt/ai-ide-init-template/codebuddy/.codebuddy/skills/delivery-orchestration Skill
3★ repo将明确需求组织为七类职责的交付步骤、并行关系、交接物与门禁。用于复杂跨角色任务。
ligyDt/ai-ide-init-template/codebuddy/.codebuddy/skills/prompt-template-library Skill
3★ repo为重复研发任务沉淀标准化提示模板。用于需求澄清、代码审查、测试生成、发布准备、风险审查、PR 描述或跨代理交接需要统一输入格式时。
ligyDt/ai-ide-init-template/codex/.agents/skills/delivery-orchestration Skill
3★ repo将已确认的需求组织为跨角色交付路径。用于需要产品、设计体验、前端、后端、质量、运维或安全职责协同,明确任务顺序、依赖、交接产物和完成门禁时。
ligyDt/ai-ide-init-template/codex/.agents/skills/prompt-template-library Skill
3★ repo为重复研发任务沉淀标准化提示模板。用于需求澄清、代码审查、测试生成、发布准备、风险审查、PR 描述或跨代理交接需要统一输入格式时。
ligyDt/ai-ide-init-template/cursor/.cursor/skills/delivery-orchestration Skill
3★ repo将已确认需求拆分为产品、体验、前端、后端、质量、运维和安全的交付顺序及交接物。用于复杂交付协同时。
ligyDt/ai-ide-init-template/cursor/.cursor/skills/prompt-template-library Skill
3★ repo为重复研发任务沉淀标准化提示模板。用于需求澄清、代码审查、测试生成、发布准备、风险审查、PR 描述或跨代理交接需要统一输入格式时。
ligyDt/ai-ide-init-template/trae/.agents/skills/delivery-orchestration Skill
3★ repo将已明确需求拆分为产品、体验、前端、后端、质量、运维和安全协作步骤。用于跨职责交付安排。
ligyDt/ai-ide-init-template/trae/.agents/skills/prompt-template-library Skill
3★ repo为重复研发任务沉淀标准化提示模板。用于需求澄清、代码审查、测试生成、发布准备、风险审查、PR 描述或跨代理交接需要统一输入格式时。
tayyabexe/skills/skills/hugging-face-datasets Skill
3★ repoCreate and manage datasets on Hugging Face Hub. Supports initializing repos, defining configs/system prompts, streaming row updates, and SQL-based dataset querying/transformation. Designed to work alongside HF MCP server for comprehensive dataset workflows.
The same file is in 1 other repository
tmusser/ai-engineering-skills/skills/constitution-lite Skill
3★ repoDefine a tiny set of stable, non-negotiable agent authority boundaries for repeated or delegated work without duplicating CLAUDE.md, AGENTS.md, or task specs.
tmusser/ai-engineering-skills/skills/handoff Skill
3★ repoCompress project context into HANDOFF.md with workflow state, analysis checkpoint, next gate, verification, current hypothesis, freshness anchors, optional gotcha references, and a resume packet for the next agent session.
Syo-M/fable-frontend-skills/.claude/skills/codex-imagegen Skill
3★ repoGenerate or AI-edit images by delegating to the Codex CLI ($imagegen / gpt-image) — sample banners, hero visuals, demo assets, image-to-image edits. Use whenever a task needs image GENERATION or AI photo editing (Claude Code cannot generate images itself). 日本語の依頼例:「画像を生成して」「サンプル画像を作って」「この写真を加工して」「バナー/ヒーロー画像が必要」。NOT for resizing/compressing existing assets (use `images-media`).
Syo-M/fable-frontend-skills/plugin/skills/codex-imagegen Skill
3★ repoGenerate or AI-edit images by delegating to the Codex CLI ($imagegen / gpt-image) — sample banners, hero visuals, demo assets, image-to-image edits. Use whenever a task needs image GENERATION or AI photo editing (Claude Code cannot generate images itself). 日本語の依頼例:「画像を生成して」「サンプル画像を作って」「この写真を加工して」「バナー/ヒーロー画像が必要」。NOT for resizing/compressing existing assets (use `images-media`).
ravi2799/ai-agent-skills/skills/eval-designer Skill
3★ repoUse this skill when designing evaluation frameworks, writing test cases, or analyzing results for AI agent systems. Triggers include "evaluate this agent", "how do I test my agent", "build an eval", "measure agent quality", "create test cases", "is my agent working well", "benchmark this prompt", or any task involving agent evaluation, metrics, grading rubrics, or performance measurement.
ravi2799/ai-agent-skills/skills/langgraph-orchestrator Skill
3★ repoUse this skill when building stateful agent workflows with LangGraph. Triggers include "use LangGraph", "build a graph workflow", "add human-in-the-loop", "create a state machine agent", "chain agents together", "build a multi-agent graph", "add checkpointing", "create agent with memory", "build a deep agent", or any task involving stateful orchestration, cycles, conditional routing, subgraphs, or complex agent pipelines beyond simple ReAct agents.
ravi2799/ai-agent-skills/skills/model-gateway Skill
3★ repoUse this skill when configuring LLM API calls through model gateways and routers. Triggers include "call an LLM", "set up OpenRouter", "configure model gateway", "add LLM to my agent", "switch models", "set up LLM factory", "create LLM client", "use LangChain with OpenRouter", or any task involving LLM API configuration, model selection, gateway routing, streaming, or provider-agnostic LLM access.
ravi2799/ai-agent-skills/skills/prompt-engineer Skill
3★ repoUse this skill whenever an agent needs to create, modify, rewrite, evaluate, or review any prompt. Triggers include "write a prompt for...", "create a system prompt", "update this prompt", "improve this prompt", "rewrite the system prompt", "optimize this prompt", "change this prompt to...", "make this prompt better", "fix this prompt", "review this prompt", "is this a good prompt", "evaluate this prompt", or any task where the agent is about to write or edit a prompt string in code, config, or instructions. This skill enforces proven prompting best practices as mandatory rules for all prompt operations.
ravi2799/ai-agent-skills/skills/skill-creator Skill
3★ repoUse this skill when writing new SKILL.md files, reviewing existing skills, or preparing skills for distribution. Triggers include "create a new skill", "write a SKILL.md", "review this skill", "publish this skill", "is this skill well-written", "make a skill for...", or any task involving authoring, evaluating, or distributing agent skill files.
ravi2799/ai-agent-skills/skills/tool-designer Skill
3★ repoUse this skill when creating, reviewing, or converting tool and function definitions for AI agents. Triggers include "write a tool definition", "create a function schema", "review this tool", "convert MCP tool", "improve tool description", "design agent tools", "build a LangChain tool", "add capability to my agent", or any task involving JSON Schema tool definitions, function calling schemas, MCP tool specs, or LangChain tools.
Sameeh07/AGENT-SKILLS/skills/LangGraph Skill
no license3★ repoLangGraph skill for building stateful agent graphs, subgraphs, checkpoints, conditional flows, parallel execution, streaming, and multi-agent orchestration.
The same file is in 2 other repositories
Sameeh07/AGENT-SKILLS/skills/deep-agents-memory Skill
no license3★ repoINVOKE THIS SKILL when your Deep Agent needs memory, persistence, or filesystem access. Covers StateBackend (ephemeral), StoreBackend (persistent), FilesystemMiddleware, and CompositeBackend for routing.
The same file is in 33 other repositories
Sameeh07/AGENT-SKILLS/skills/deep-agents-orchestration Skill
no license3★ repoINVOKE THIS SKILL when using subagents, task planning, or human approval in Deep Agents. Covers SubAgentMiddleware, TodoList for planning, and HITL interrupts.
The same file is in 34 other repositories
Sameeh07/AGENT-SKILLS/skills/langchain-rag Skill
no license3★ repoINVOKE THIS SKILL when building ANY retrieval-augmented generation (RAG) system. Covers document loaders, RecursiveCharacterTextSplitter, embeddings (OpenAI), and vector stores (Chroma, FAISS, Pinecone).
The same file is in 35 other repositories
antohins/seo-tools-mcp/io.github.antohins/seo-tools-mcp-wordstat MCP server
3★ repoMCP server for Yandex Wordstat keyword frequencies (official API v2). Free Yandex Cloud Search API.
SEO Tools: XMLStock SERP (Google + Yandex)
antohins/seo-tools-mcp/io.github.antohins/seo-tools-mcp-xmlstock MCP server
3★ repoMCP server for Google & Yandex SERP via XMLStock (read-only).
antohins/seo-tools-mcp/io.github.antohins/seo-tools-mcp-ywm MCP server
3★ repoMCP server for Yandex.Webmaster search queries (read-only).
jposluns/grc_library/dev-security/claude-rules/skills/skill-authoring-discipline Skill
2★ repoWhen adding a new skill to the governance pack's `skills/` collection, or when substantively revising an existing skill's structure or frontmatter, apply the pack's established eight-section structural template (YAML frontmatter with `derives_from`, Overview, When-to-Use, Process, Red Flags, Verification, Common Rationalizations, See Also) and validate trigger-accuracy with representative positive, negative, and boundary prompts. Catches structural drift across pack skills that accumulates silently as the pack grows: section compression, missing cross-references, missing rationalizations table, vague description wording. The mechanical derives-from reference gate only verifies `derives_from` resolution; everything else relies on authorial discipline that this skill codifies.
Ed3Design/ed3design-engineering-bundles/cad-design/skills/fusion-mcp-bridge Skill
2★ repoUse when driving Fusion 360 from Claude Code through the local MCP-Bridge (HTTP server on 127.0.0.1:7654 with bearer-token auth, single-threaded Custom-Event handler). Covers the API quirks that consistently bite without a documented protocol: component activation before geometry, single-revolve over loft+combine-join, cut-before-shell sequencing, tool-body combine-cut for complex cuts, boolean-union before cuts on multi-body components, bridge timeout/lock recovery, screenshot setup, and the `fusion` CLI that replaces curl boilerplate. Trigger on phrases like "Fusion 360 via MCP", "fusion-mcp-bridge", "Fusion Custom-Event", "run Fusion API script", or any `fusion` CLI command. Do NOT load for general parametric CAD (use `cad-construction`), mechanical-design rules (use `mechanical-design-principles`), or OpenSCAD/FreeCAD (use `cad-api-scripting`).
LLl0k0laD/agent-skills/skills/context-engineering Skill
2★ repoUse when starting a new session, when agent output quality degrades, when switching between tasks, or when you need to configure rules files and context for a project.
LLl0k0laD/agent-skills/skills/using-agent-skills Skill
2★ repoUse when starting a session or when you need to discover which skill applies to the current task. This is the meta-skill that governs how all other skills are discovered and invoked.
Paretofilm/superpowers-gstack/skills/context-handoff Skill
no license2★ repoWrite a structured handoff file (YAML frontmatter + prose) to docs/superpowers/handoff.md before /clear or /compact. Different from /context-save (gstack) — survives across machines and restores without gstack installed.
couzip/cc-forever/skills/index-memory Skill
2★ repoSave conversation Q&A pairs to persistent memory for later retrieval
couzip/cc-forever/skills/search-memory Skill
2★ repoSearch past conversations using semantic search. Use when user asks about previous discussions or wants to recall past information.
voxpelli/claude-beads/skills/swarm-wave Skill
2★ repoOrchestrate multi-agent development sprints with wave-based parallelism. Use when the user wants to plan a swarm sprint, partition work into file-disjoint waves, map file contention across open issues, run a post-wave quality gate with review agents, manage agent backpressure, run a parallel research wave, or coordinate multiple concurrent agents on a shared codebase. Works with or without beads: sources waves from the beads backlog, from a ROADMAP.md, or from a manually supplied work list. Trigger phrases: 'swarm sprint', 'wave plan', 'launch wave', 'execute wave', 'post-wave gate', 'contention map', 'research wave', 'parallel agents', 'multi-agent sprint', 'agent wave', 'swarm orchestration', 'swarm from ROADMAP', 'wave plan without beads', 'swarm with manual list'.
ForgeyClap/claude-forge/.claude/skills/forge-rag Skill
2★ repoForge playbook for AI chatbots and RAG systems. Use when building a chatbot, assistant, RAG, or knowledge-base Q&A over documents — keywords: chatbot, RAG, retrieval, embeddings, vector search, ingestion, system prompt, knowledge base, assistant, hallucination, source-aware, lead capture. Covers retrieval logic, grounding, fallback behavior, and safety.
ForgeyClap/claude-forge/plugins/forge/skills/forge-agent-report Skill
2★ repoStructured completion-report contract every dispatched Boss ends its final message with, plus the forge-bin/forge-report.cjs tool that parses, validates, and ingests it. Use when dispatching any subagent (include the contract requirement in the prompt) and when a dispatched agent's final message comes back (validate/ingest it before treating the work as done).
ForgeyClap/claude-forge/plugins/forge/skills/forge-rag Skill
2★ repoForge playbook for AI chatbots and RAG systems. Use when building a chatbot, assistant, RAG, or knowledge-base Q&A over documents — keywords: chatbot, RAG, retrieval, embeddings, vector search, ingestion, system prompt, knowledge base, assistant, hallucination, source-aware, lead capture. Covers retrieval logic, grounding, fallback behavior, and safety.
fernando-bertholdo/4-successful-AI-life/plugins/prompt-master/upstream Skill
2★ repoGenerates optimized prompts for AI tools. Activates only when the user explicitly asks to write, fix, improve, or adapt a prompt for a specific AI tool (LLM, Cursor, Midjourney, image AI, video AI, coding agents, etc.). Does not activate for general conversation, coding tasks, document writing, or other non-prompt-engineering work.
The same file is in 3 other repositories
RockyHong/super-bootstrap/plugins/super-bootstrap/skills/log Skill
2★ repoCapture front door. Logs 1..N observations as card files at docs/work/{ID}.md (BUG / DEBT / GAP), classified + deduped on Sonnet. A duplicate carrying a new fact lands as an Amendment block on the owning card (amended branch). Use when the user says "log this", "track that", "note this down", or types `/super-bootstrap:log <observation>` — and when Claude needs to file its own findings (out-of-scope findings from a review, audit, or returned subagent report). Dispatches the `log` subagent so classify + write run off the gateway model. Captures raw — the real/worth/now call is triage's at `/super-bootstrap:todo` pickup. Does NOT triage (that is the `/super-bootstrap:todo` triage lane). Feature ideas log as GAP.
RockyHong/super-bootstrap/plugins/super-bootstrap/skills/triage-report Skill
2★ repoDrain the `.review/` report queue. Resolves un-triaged scan/audit reports (producer-agnostic — check-docs-consistency and any other scanner), dispatches the `triage-report` subagent (Sonnet) one report at a time for per-finding dispositions (promote / patch / dup / needs-investigation / dismiss), then the gateway absorbs: promotes batch through /super-bootstrap:log, doc-mechanical patches land per CLAUDE.md § Dispatch, investigations route to /super-bootstrap:triage, report deleted at close-out once every finding holds a terminal verdict. Use when a report lands in `.review/` or the user asks to process review findings. Does NOT fix findings.
spideynolove/claude-code-in-action/00-materials/claude-howto/03-skills/claude-md Skill
2★ repoCreate or update CLAUDE.md files following best practices for optimal AI agent onboarding
spideynolove/claude-code-in-action/05-subagents/.claude/skills/mcp-knowledge-graph Skill
2★ repoThe process I experienced firsthand through the https://anthropic.skilljar.com/claude-code-in-action course using real-world daily projects
spideynolove/claude-code-in-action/27-tdd-conductor-llmdev/.claude/skills/embedding-strategies Skill
2★ repoSelect and optimize embedding models for semantic search and RAG applications. Use when choosing embedding models, implementing chunking strategies, or optimizing embedding quality for specific domains.
The same file is in 2 other repositories
spideynolove/claude-code-in-action/27-tdd-conductor-llmdev/.claude/skills/llm-evaluation Skill
2★ repoImplement comprehensive evaluation strategies for LLM applications using automated metrics, human feedback, and benchmarking. Use when testing LLM performance, measuring AI application quality, or establishing evaluation frameworks.
The same file is in 2 other repositories
spideynolove/claude-code-in-action/27-tdd-conductor-llmdev/.claude/skills/prompt-engineering-patterns Skill
2★ repoMaster advanced prompt engineering techniques to maximize LLM performance, reliability, and controllability in production. Use when optimizing prompts, improving LLM outputs, or designing production prompt templates.
The same file is in 2 other repositories
spideynolove/claude-code-in-action/27-tdd-conductor-llmdev/.claude/skills/rag-implementation Skill
2★ repoBuild Retrieval-Augmented Generation (RAG) systems for LLM applications with vector databases and semantic search. Use when implementing knowledge-grounded AI, building document Q&A systems, or integrating LLMs with external knowledge bases.
The same file is in 1 other repository
spideynolove/claude-code-in-action/28-continuous-patterns/.claude/skills/create-handoff Skill
2★ repoCreate handoff document for transferring work to another session
The same file is in 1 other repository
spideynolove/claude-code-in-action/29-skill-authoring/.claude/skills/agentic-engineering Skill
2★ repoOperate as an agentic engineer using eval-first execution, decomposition, and cost-aware model routing.
The same file is in 10 other repositories
spideynolove/claude-code-in-action/29-skill-authoring/.claude/skills/skill-security-auditor Skill
2★ repoSecurity audit and vulnerability scanner for AI agent skills before installation. Use when: (1) evaluating a skill from an untrusted source, (2) auditing a skill directory or git repo URL for malicious code, (3) pre-install security gate for Claude Code plugins, OpenClaw skills, or Codex skills, (4) scanning Python scripts for dangerous patterns like os.system, eval, subprocess, network exfiltration, (5) detecting prompt injection in SKILL.md files, (6) checking dependency supply chain risks, (7) verifying file system access stays within skill boundaries. Triggers: "audit this skill", "is this skill safe", "scan skill for security", "check skill before install", "skill security check", "skill vulnerability scan".
The same file is in 1 other repository
atuljha23/holocron/skills/context-engineering Skill
2★ repoApply the Write/Select/Compress/Isolate framework to manage the context window. Use when the task is long, the transcript is bloated, the model is drifting, or before /holocron:handoff.
atuljha23/holocron/skills/token-economy Skill
2★ repoRubric for keeping Claude Code sessions cheap without losing quality — model tiering, cache hygiene, Read discipline, MCP pruning, subagent isolation. Use when cost feels wrong, before setting a budget, or when onboarding a teammate.
batterfried-philosophy172/Agent-Startup-Skills/skills/saas-accelerator Skill
2★ repoMaster orchestrator skill that coordinates a full multi-agent startup pipeline. Triggers on: build SaaS, launch startup, startup accelerator, build product, create MVP, SaaS builder, million dollar startup, build me a business, or using the Startup Factory. Coordinates 8 specialized sub-skills in a gated, human-approved sequence.
renalpelvisdeinocheirus70/ozor-skills/skills/ozor-prompt-optimizer Skill
2★ repoAnalyze and improve existing Ozor.ai video prompts for better output quality. Use this skill whenever the user has an Ozor prompt that produced mediocre results, wants to improve a video prompt, needs help debugging why a video didn't turn out well, or wants expert feedback on their prompt before running it. Trigger on phrases like 'improve this prompt', 'optimize my Ozor prompt', 'why did my video look bad', 'fix this prompt', 'make this prompt better', 'review my Ozor prompt', 'the video didn't turn out right', 'prompt feedback', 'better results from Ozor', or any request to refine, debug, or enhance an existing Ozor video prompt.
The same file is in 1 other repository
Basic-XYZ/baku-skills/baku-skill-manager Skill
2★ repo管理全局 Agent Skill 的安装、认领、软链接、更新、三方合并、冲突恢复、卸载和 Codex 使用记录同步。用户说“安装这个 skill”“检查 skill 更新”“把 skill 挂到 Claude/Codex”“更新但保留本地修改”“同步 Codex skill 使用记录”或“卸载 skill”时使用。
jperrello/C0BALT_CUT/.claude/skills/director-pass Skill
no license2★ repoAn agentic, vision-driven final QA/repair pass — a Claude "director" WATCHES the finished short (a contact sheet of frames sampled across the whole clip + the transcript + the sidecar plans) and decides, in natural language, what is broken ANYWHERE (cold open, dead/rambling tail, a tonally-wrong b-roll match, a wrong-person punch-in, mistimed captions, a too-hot music bed, a flat hook) — not just the cold open, not just a fixed route vocabulary. It then APPLIES the fixes it can from a bounded, pixel-safe set (tail_trim via cut-clip; music_down via a bg-music re-mix) and SURFACES everything else as an honest edit list, re-reviewing up to DIRECTOR_MAX_ITERS times. Layered ON TOP of grade-clip (the cheap proxy floor) + fix-cold-open (the closed-vocab cold-open repair). NON-FATAL, idempotent (.dpmeta). Two modes — preventive in-chain (the pre-mix .ctaed.mp4 + clip_NN.* sidecars are co-located, so music_down can re-mix) and curative standalone/backlog (a finished output clip → tail_trim only, everything else surfaced — never fabricate a baked-in repair).
multiplex-ai/muggle-ai-teams/skills/agent-harness-construction Skill
2★ repoDesign and optimize AI agent action spaces, tool definitions, and observation formatting for higher completion rates.
The same file is in 7 other repositories
multiplex-ai/muggle-ai-teams/skills/agentic-engineering Skill
2★ repoOperate as an agentic engineer using eval-first execution, decomposition, and cost-aware model routing.
The same file is in 10 other repositories