Context & AI engineering
25,864 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
ekreloff/claude-self-eval/skills/self-eval Skill
no license0★ repoHonestly evaluate AI work quality using a two-axis scoring system. Use after completing a task, code review, or work session to get an unbiased assessment of what was accomplished. Detects score inflation, forces devil's advocate reasoning.
arterialist/mcphero-skills/skills/create-mcp-server Skill
no license0★ repoCreate, deploy, and manage MCP (Model Context Protocol) servers using the MCPHero platform via the mcpheroctl CLI. Use this skill when the user wants to build an MCP server, deploy tools that wrap APIs or databases, automate MCP server creation, or connect AI clients (Claude Desktop, Cursor, etc.) to custom tools through MCPHero.
AmrMohamad/figma-to-ios-ui/figma-mcp Skill
no license0★ repoAuthoritative workflow for robust Figma MCP usage across Codex and Claude Code. Use for any task involving Figma design extraction, FigJam parsing, node IDs, design tokens, screenshot-based validation, design-to-code, or prompts containing a Figma URL. Trigger on phrases like "use Figma MCP", "get design context", "get metadata", "extract variables", "implement this Figma", or "compare with Figma screenshot". Always load this skill first before framework-specific implementation skills. For iOS output, load `figma-to-ios-ui` after this skill.
arendon1/agent-skills/domain/analyze-model Skill
0★ repoFetches live model data from OpenRouter and Artificial Analysis APIs, merges benchmarks with pricing, runs deep cost analysis on usage logs, exports usage from LLM surfaces (OpenCode, Pi, VS Code, etc.), analyzes subscription efficiency, and forecasts future spend with cheaper-model alternatives. Use when comparing LLM costs or quality, analyzing total API spend from usage logs, exporting agent usage for cost analysis, evaluating subscription cost efficiency, forecasting future LLM costs, finding cheaper alternatives, or when the user asks about model pricing, intelligence index, speed, or token costs across providers.
Panno-ai/skills/skills/llm-resource-usage Skill
0★ repoAnalyzes the user's LLM/AI usage across all local harnesses and coding agents (Claude Code, OpenClaw, OpenCode, Goose, Codex CLI, Gemini CLI, Cursor, Amp, etc.), then enriches that data with environmental impact (energy, CO₂, water), likely datacenter locations, and fun equivalency comparisons — and produces a visual infographic + narrative report. Use this skill whenever the user asks about: their AI usage, token consumption, LLM costs, energy footprint of AI, how much they've spent on AI, carbon footprint of their coding agent, usage stats, "how many tokens have I used", AI environmental impact, "what's my AI bill", or wants to see charts/visuals of their LLM activity. Also trigger when they ask to "track", "analyze", "visualize", or "report" on AI or token usage.
hereandnowai/agent-skills/deepagents Skill
0★ repoComprehensive guide for building Python apps with the LangChain `deepagents` library — an agent harness with built-in planning, virtual filesystem, sub-agents, human-in-the-loop, MCP integration, and pluggable backends for long-running, context-heavy tasks. Use whenever the user writes code with `deepagents`, calls `create_deep_agent` or `async_create_deep_agent`, or builds a research agent, coding agent, data-analysis agent, or multi-step LangChain/LangGraph agent that needs planning + sub-agents + virtual filesystem + context engineering. Trigger on any mention of "deepagents", "deep agent", "deep agents", "LangChain agent harness", "create_deep_agent", "agent with planning tool", "multi-agent orchestration with subagents", or when the user wants an agent that handles long-horizon tasks via context offloading, summarization, and isolation. Also use when migrating older 0.0.x code to the 0.4+ API. Prefer over generic LangChain guidance whenever `deepagents` is the framework.
ph3on1x/adveloop/skills/adveloop Skill
0★ repoRun a GAN-inspired adversarial development loop — the Planner drives a Generator and Evaluator in fresh cmux panes, gated by hard pass/fail per deliverable, adapted from Anthropic's harness-design guidance for long-running agentic apps. Use when the user asks to /adveloop, run an adversarial dev loop, spawn Planner-Generator-Evaluator panes, build with adversarial verification, or audit/harden existing code with a skeptical evaluator. Requires cmux and the claude-cmux-skill:cmux skill.
Claude code environment backup
gaoguobin/codex-environment-backup/skills/claude-code-environment-backup Skill
0★ repoClaude Code environment backup, restore, doctor, and list workflows for local ~/.claude state. Use for requests like back up Claude Code, restore Claude Code backup, list/check Claude Code backups, or install/update/uninstall the backup skill.
Lukk17/agent-standards/.agents/skills/agentic-engineering Skill
0★ repoOperate as an agentic engineer using eval-first execution, decomposition, and cost-aware model routing.
0SxD/notebooklm-mcp-skill Skill
0★Query, inspect, and manage Google NotebookLM notebooks from Claude Code. Trigger whenever the user says "query notebook", "ask notebooklm", "search notebooks", "list notebooks", "notebooklm server info", "get notebook", "inject source into notebook", "what did notebooklm say about", or any request to pull evidence from a named NotebookLM notebook as grounding for an artifact. Covers three access paths: (1) the mcp__notebooklm-mcp__* MCP tool surface (35 tools) when the stdio server is loaded, (2) a Claude-in-Chrome browser route against notebooklm.google.com when the MCP is unavailable, (3) a Playwright fallback using notebooklm_uploader.py for bulk source injection. Always ask the user which notebook to query before any call. Never auto-select. Never fabricate notebook IDs or contents. Every query runs through the 100 Percent Confidence Loop and every synthesized answer passes the Trinity Dialectic gate.
Shirolin/i18n-agent-skill Skill
0★High-performance frontend internationalization expert. Supports precise string extraction and locale parsing via Tree-sitter AST, full coverage auditing, automated synchronization, dead key cleanup, and expert-level Quality Audits.
damienen/basd-coding-dispatch/skills/basd-coding-dispatch Skill
0★ repoHermes/OpenClaw-first dispatch workflow for approval-gated AI coding from Telegram and mobile chat. Use when planning, implementing, reviewing, or coordinating Codex, Claude, split-worker, subagent, or generic agent workflows.
0SxD/advisor-trackable-eval-skills-v01/skills/eval_criteria_create Skill
0★ repodavdittrich/planning-with-beads Skill
no license0★Use when writing any ticket, issue, PR description, or bug report. Mandatory when starting complex task (3+ steps), research, multi-session work, or any multi-agent handoff.
AIEngineerX/printr-agent-skills/printr-agent-payments Skill
0★ repoUse when building a Solana paywall — charging users for agent actions with on-chain memo-match invoice verification. Generates cryptographically-random memos, builds SOL or USDC transfer transactions for client signing, and verifies payments on-chain by scanning the treasury wallet's signature history. Platform-agnostic — works for any Solana SPL mint and any Solana wallet adapter. Pairs with `printr-swap` and `printr-tokenized-agent` for a full agent-revenue + buyback-and-burn loop on Printr POB tokens.
Scribeless agent direct mail workflows
scribeless/skills/skills/scribeless-agent-direct-mail-workflows Skill
0★ repoDesign and review AI-assisted, API-led Scribeless direct mail workflows. Use when a user wants an AI agent, LLM workflow, custom app, server-side automation, CRM automation, data pipeline, or generated creative system to create recipients, generate personalized variables or HTML front/back content, validate payloads, trigger previews, and safely operate direct mail campaigns through the Scribeless API.
Alpha-Park/genpark-accio-csuite-manager Skill
no license0★Multi-Agent C-Suite Orchestrator. Manages a team of specialized AI agents (CMO, CTO, CFO, etc.) for cohesive business strategy.
The same file is in 1 other repository
youlihui12-ux/agent-skill-monetization-kit/examples/minimal-skill Skill
no license0★ repoUse when you need a tiny example of a packaged agent skill.
0SxD/notebooklm-research-protocol/skills/notebooklm-orchestrator Skill
0★ repoTeaches the agent how to interact with NotebookLM via MCP, sanitize output via the ZCR pattern, maintain strict notebook selection compliance, and log all research via stigmergy.
Genpark nano banana video orchestrator
Alpha-Park/genpark-nano-banana-video-orchestrator Skill
no license0★Precision Video Orchestrator. Performs precise object/character masking and replacement in motion video using advanced AI vision workflows.
The same file is in 1 other repository
0SxD/github-curator-v01/github-curator-v0.1.0/skills/github-curator Skill
no license0★ repoUse this skill when curating, evolving, or publishing a GitHub repository that contains Anthropic Agent Skills, especially when the repo must remain self-consistent across edits. Triggers: any request to add a skill, edit a skill, bump a version, cut a release, audit the repo against the Agent Skills spec at agentskills.io, validate AGENTS.md and CLAUDE.md alignment, run the four-gate publish-readiness check on this repo itself, or onboard a new contributor to the curator pattern. The skill operates on its own repository as the first example: every action it instructs the agent to take, it has already taken on this repo. Do NOT use this skill for arbitrary GitHub tasks unrelated to skill curation, for non-Skills repos, or when the user has asked for code-level changes to a separate codebase.
danzimon-rgb/triangulate/skills/triangulate Skill
0★ repoUse BEFORE shipping high-stakes output, considering an architectural alternative, working with compliance/legal/safety-critical content, or stuck in a debug loop >2 cycles — fork the question to an independent source of truth (different model, fresh subagent, docs) before trusting your first answer. The meta-decision skill that fires existing review skills (second-opinion, adversarial-reviewer, dispatching-parallel-agents) at the right moments. Anchoring on first instinct is a known LLM failure mode; the cost of a 30-second outside review is far lower than the cost of shipping the wrong pattern.
rema424/ai/skills/github-issue-operations Skill
0★ repoUse when operating GitHub Issues across repositories with `gh` or a GitHub MCP server. Covers issue create/edit/view/close, label inspection, bulk issue workflows, sub-issue linking, `gh api` payload typing, issue number vs REST id, auth checks, and restart-safe recovery after partial failures.
biyearly-mesothelioma790/skills/claude-session-introspect Skill
no license0★ repoInspect Claude Code session JSONL files at ~/.claude/projects/ to extract real conversation telemetry: token counts (input/output/cache reads/cache writes), assistant turn counts, human prompt counts, tool-use counts, compaction boundaries, and the contents of compaction summaries. Use this skill when the user asks "how many tokens did this session use", "how many prompts have I sent", "show me the stats for this conversation", "what got compacted", "where are the compaction boundaries", "introspect the session", "do brain surgery on the JSONL", or wants any data point that lives inside the on-disk session log rather than the live context window. Inspired by Tal Raviv's "I wanted to know how compaction works" article.
itsyuvallavi/monochrome-particle/skills/monochrome-particle Skill
0★ repoBuild or port a configurable full-viewport Three.js layered particle wave background using Points, ShaderMaterial, GLSL waves, soft sprite textures, additive blending, and mobile/desktop density tiers. Use for Next.js/React canvas backgrounds, Three.js Points shaders, monochrome dots, particle wave, animated hero backdrop, or requests to change particle colors, speed, direction, density, point size, opacity, or camera zoom. Portable Markdown instructions: applies in any coding agent or IDE that loads SKILL.md (Cursor, Claude Code, OpenAI Codex, Google Gemini, Windsurf, JetBrains AI, Copilot Chat, VS Code–based agents, or paste into web UIs). Does NOT cover page transitions, SandstormProvider, sandstorm hooks, button-triggered navigation, or route orchestration; only the static/ambient background.
SoulFaro/agent-first-codebase/skills/agent-first-codebase Skill
0★ repoDiagnose and transform a codebase so coding agents (Claude Code, Codex, and similar) can work in it reliably and autonomously. Based on the practices OpenAI's Codex team used to ship a 1M-line internal product with zero human-written code. Use this skill whenever the user wants to set up or restructure AGENTS.md / CLAUDE.md, organize a docs/ directory for AI agents, diagnose why Claude Code keeps making mistakes in their repo, make their codebase "AI-friendly" or "agent-first", references "harness engineering" or Codex best practices, audits an existing repo's agent-readability, or is starting a new repo they plan to develop heavily with AI assistance. Also trigger if the user complains that their agent "forgets context", "reinvents patterns", or "ignores conventions" — those are symptoms this skill addresses at the structural level.
ImL1s/claude-skills-toolkit/skills/opencli-llm-advisors Skill
0★ repoUse opencli to drive LLM chat apps (Gemini web, ChatGPT Desktop) as external advisors — bypasses API 429/quota by using the user's logged-in browser/desktop-app session. Triggers when: Gemini CLI returns 429 or capacity-exhausted, user wants a CCG advisor and API is unavailable, user wants to use their ChatGPT Plus subscription instead of API, or user says 'ask Gemini/ChatGPT via opencli'. Covers `/u/N` account switching, clicking the in-page mode picker to select highest-tier model (Gemini Pro 3.1), ChatGPT Desktop native control, and common failure modes. Does NOT apply for: normal API calls that work, image generation (use `opencli chatgpt image` directly), or headless/CI workflows.
Ben2W/delegate-ui-to-claude/skills/delegate-ui-to-claude Skill
no license0★ repoUse Claude Code as a headless frontend design refinement pass after Codex scaffolds or implements UI code. Use whenever Codex is asked to create, modify, polish, redesign, review, or verify user-facing frontend code, visual design, layout, styling, responsive behavior, accessibility, components, pages, or interactions; Codex may write the functional frontend implementation, but must ask Claude Code to rewrite/refine the design with embedded frontend-design instructions before finalizing user-facing UI work.
joshuaporth/appsec-skill/.cursor/skills/benchmark Skill
0★ repoRuns the full 30-challenge secure-code-review benchmark end-to-end via scripts: Findings (parallel across 01-30), then Scoring (parallel across 01-30), then scoreboard generation from frozen scoring artifacts. Invoke from Cursor Agent with /benchmark.
gustmrg/ai-workflows/skills/devlog Skill
no license0★ repoUse the DevLog CLI to record development work after an agent coding session, including agent-written concise, detailed, formal, or impersonal log entries without relying on DevLog's --ai flag. Use this skill whenever the user asks to log work, write a devlog entry, summarize what just changed, capture today’s coding session, record progress for a project, choose a DevLog style/template, or update their daily development notes from the current OpenCode/Codex/Claude Code context. This skill is especially relevant after making code changes, running tests, debugging issues, reviewing code, or finishing a task in a coding harness.
lyoneel/ly-agent-skills/agent Skill
0★ repoLoad and activate a custom agent from the agents folder. Use when you need to switch to a specialized agent definition for focused task execution.
sanketsao/prd-for-agents Skill
no license0★Use this skill whenever an agent or user asks to write, draft, create, or produce a PRD, product requirements document, product spec, or feature spec — especially when the output will be consumed by an AI coding agent, developer agent, or multi-agent pipeline rather than a human PM. Also trigger when asked to "write requirements for X", "spec out X", "define what needs to be built for X", or "create a handoff doc for Developer". This skill produces agent-optimized PRDs that include quantitative success metrics, user stories, open questions, risks, dependencies, and a sequenced build order that a Developer agent can execute directly without clarification. Always use this skill in preference to writing a PRD from scratch — it enforces completeness gaps that generic PRDs miss.
shihabshahrier/softco/skill Skill
0★ repoBuild, deploy, and ship any software product end-to-end as a complete autonomous software company inside Claude Code. Triggers on: "build me a...", "create an app/API/SaaS", "design the architecture for...", "set up CI/CD", "dockerize", "kubernetes", "terraform", "integrate LLM/AI/GPT/Claude/Gemini/ RAG/agent", "deploy to AWS/GCP/Azure/Fly/Render", "review my code", "audit this repo", "fix this bug", "scaffold a project", or any open-ended request to create, extend, or ship software. SoftCo acts as PM, architect, senior engineer, DevOps, QA, security engineer, and tech writer. Docs and tests before code. Do not invoke for one-line edits, trivial single-file questions, or pure conceptual Q&A that needs no product output.
guynhsichngeodiec/cc-skills-golang/skills/golang-benchmark Skill
0★ repoGolang benchmarking, profiling, and performance measurement. Use when writing, running, or comparing Go benchmarks, profiling hot paths with pprof, interpreting CPU/memory/trace profiles, analyzing results with benchstat, setting up CI benchmark regression detection, or investigating production performance with Prometheus runtime metrics. Also use when the developer needs deep analysis on a specific performance indicator - this skill provides the measurement methodology, while golang-performance provides the optimization patterns.
pccly/project-indexer Skill
0★Use when the user asks to "index this project", "map the codebase", "generate a project index", "create PROJECT_INDEX.md", "refresh/update the project index", or wants any AI coding agent (Claude Code, Cursor, Codex, Gemini CLI, Windsurf, Copilot, Cline, etc.) to understand the codebase layout without using Glob/Grep every turn. Also use when navigating an unfamiliar repo where file search is expensive and a structured index would cut tool calls. Always trigger when the user mentions indexing, cataloging, summarizing, or documenting a project's file structure — even phrased casually.
Sepi0l-brd/ai-native-foundation Skill
0★Use when the user wants to set up the "AI Native Foundation" framework on their Claude Code - positions AI as a senior tech lead so juniors and PMs ship production-grade code. Installs CLAUDE.md (operational rules), hooks (security guardrails), architectural skills (webhook design, identity coordination, atomic transactions, decision categorization, AI collaboration), memory conventions, working-style manifesto, senior traps catalog, and optional Obsidian vault. Triggered by phrases like "set up AI Native Foundation", "install this framework", or when the user references this repo by name.
fermeridamagni/skills/.agents/skills/skill-creator Skill
0★ repoCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
The same file is in 66 other repositories
Montesplenic117/CC/src/skills/bundled/claude-api Skill
no license0★ repoRebuild Claude Code CLI source from source maps into runnable TypeScript for study, debugging, and local development
okayus/okayus-skills/skills/claude-code-docker-sandbox Skill
0★ repoSet up a Docker Compose-isolated development environment where dependency installs and agent-run code execute behind a default-deny network firewall, never touching the host. Use when you want to run `npm install` / `cargo build` / `cabal build`, untrusted dependencies, or Claude Code itself in a container to contain supply-chain attacks (npm postinstall, Rust build.rs, Haskell Setup.hs) — without depending on VS Code. Reuses Anthropic's published devcontainer Dockerfile + init-firewall.sh, driven by a plain docker-compose.yml so any host editor works via bind mount; Rust/Haskell toolchains are opt-in build args on the node:20 base. Covers the egress allowlist (fatal vs OPTIONAL non-fatal domains), build-time-vs-runtime network, keeping Claude Code current inside the firewall (native updater's host is blocked — update via npm at start), why the /model picker hides flag-gated models until DISABLE_TELEMETRY is unset, bypassPermissions-by-default (container-scope, not repo-shared), and host-side git hygiene.
SoulFaro/copilot-cloud-agent/skills/copilot-cloud-agent Skill
0★ repoCreate, assign, and track GitHub Copilot cloud coding agent tasks via gh agent-task. Handles task creation, model selection, issue assignment, and live log streaming.
peak6-labs/floop-skills/floop-mcp-tools Skill
0★ repoUse when connecting to Floop through MCP, calling Floop tools, proposing tasks without an API key, authenticating with a Floop key, discovering workers, creating tasks, approving proof, or managing task lifecycle through model tools.
himynameisben/claude-rc Skill
0★Manage Claude Code Remote Control (RC) sessions via tmux with a per-session YAML registry. Use when the user wants to start a new RC session, list existing sessions, attach to a running session, kill a session, clean up dead sessions, or mentions claude-rc, remote control session, RC session, tmux session for Claude Code, 開 rc session, 新開 session, 列出 session, attach session, 清理 session, 殺掉 session.
mj-deving/pai-skills/skills/Agents Skill
no license0★ repoCompose CUSTOM agents from Base Traits + Voice + Specialization for specialized perspectives. USE WHEN create custom agents, spin up agents, specialized agents, agent personalities, available traits, list traits, agent voices, compose agent, load agent context, agent profile, spawn parallel agents, launch agents. NOT for agent teams/swarms (use Delegation skill → TeamCreate) or Codex delegation (use Delegation skill → CodexBridge).
fbsmna-coder/karpathy-pro-max/skills/anti-hallucination Skill
0★ repoNever invent APIs, file paths, line numbers, library functions, CLI flags, or framework features. Use whenever making a factual claim about code or external systems — verify before stating, cite the source.
0SxD/claude-code-safety-hooks Skill
no license0★Two deterministic Claude Code hooks: a PreToolUse Bash command guard against destructive operations, and a PreToolUse research-output validator that requires approved-source citations on Write/Edit calls.
sekachev/codex-new-agent-skill/codex-new-agent Skill
0★ repoCreate a new adaptive Codex-native agent workspace from a reusable template. Use when the user asks to create, scaffold, initialize, bootstrap, or install a personal/project agent with onboarding, durable memory, daily logs, Codex hooks, and dreaming-style memory consolidation.
selinayfilizp/sunday-letter/skills/sunday-letter Skill
0★ repoWrite a grounded weekly Sunday Letter from dated local agent conversations. Use when the user asks for a Sunday Letter, a weekly reflection, what changed this week, or when the sunday-letter schedule fires. The supported reference paths are Codex local and Claude Code local. The runtime validates provenance, enforces default silence, sanitizes HTML, updates an atomic ledger, and rebuilds a private local archive.
Alpha-Park/tecdo-navos-skill Skill
no license0★Full-funnel AI marketing orchestrator powered by Tec-Do 2.0's Navos framework for autonomous global advertising campaigns and media buying.
The same file is in 1 other repository
vaibhav-khattar/prompt-optimizer-skill Skill
no license0★Use this skill whenever the user wants to improve, optimize, rewrite, or generate a better version of any prompt they are writing for Claude or another LLM. Trigger on phrases like "improve this prompt", "make this prompt better", "optimize my prompt", "rewrite this prompt", "help me write a good prompt", "how do I prompt Claude to...", "turn this into a proper prompt", "my prompt isn't working", "generate a prompt for me", or any time the user shares a rough idea or instruction and wants it turned into a high-quality Claude prompt. Also trigger when the user says "prompt engineer this", "make Claude do X better", or asks for a system prompt. Always use this skill — even for casual prompt requests — because applying the right techniques transforms mediocre prompts into consistently excellent ones.
Swarm-AI-Labs/pie-skills/skills/pie-pieui Skill
no license0★ repoUse for any task in the Pie ecosystem — the Python `pie` CLI and FastAPI backend, the TypeScript `pieui` CLI and Next.js/React frontend, or both together. Covers project scaffolding, cards and pages, ajax handlers and the `stored` prop, backend↔frontend metadata sync, the card preview harness and pie-show MCP server, Centrifuge realtime, MongoDB/Beanie, Cloudflare Workers, the Express backend, remote component storage, and PieUI style isolation.
SizovOleg/project-dna/skills/architect-cc-workflow Skill
0★ repoUniversal workflow protocol для пары architect (Claude.ai) + Claude Code в DNA/RNA проектах. Используй ВСЕГДА при: формулировании commands к Claude Code, написании DevPrompt'ов, ревью workflow между sessions, обнаружении систематических ошибок architect-side, планировании phase transitions. Триггеры: «напиши команду Claude Code», «напиши DevPrompt», «как мне дать задачу», «почему architect ошибается», «как валидировать решение», «что architect решает что Claude Code», «верификация спеков», «verification protocol», «error budget», «assumptions section», «boundaries Claude Code architect». Скилл кодифицирует уроки из Phase 1-2 sat-revisit-tool: ~10% architect error rate из-за LLM structural limits, защиты через verification + probe-first + assumption declaration + state refresh + hard iteration limits. НЕ для DNA creation (project-dna) или PRD writing (prd-coauthoring) или scientific methodology (research-with-ai).
tnagatomi/skills/skills/agent-test-with-report Skill
no license0★ repoDrive the agent-browser CLI to exercise the features a pull request changes, then post an evidence-only findings report back to the PR. Depends on the agent-browser skill (https://github.com/vercel-labs/agent-browser) for the actual browser operation. Use this skill when the user wants an agent to manually test a PR in a real running app and leave a report on the PR, with phrases like "agent-test this PR", "browser-test the PR and comment results", "run exploratory testing and post a report", "check this change in the running app and report on the PR". The report collects evidence (operations, screenshots, findings, unverified scope) for a human reviewer — it never produces a pass/fail verdict.
swiftpostlabs/agentic-tools/.agents/skills/ref-sp-agents-hooks Skill
0★ repoAuthor agent lifecycle hooks that run deterministic shell commands at session, prompt, tool, and stop events across Claude Code, GitHub Copilot CLI, VS Code, and Gemini CLI. Use when: creating or editing a hook, choosing a lifecycle event, writing a hook script that reads stdin JSON and returns an allow/deny/context decision, making a hook portable across agents, or debugging why a hook does not fire or block.
Ahaaad3759/PromptSkill4image Skill
0★Universal prompt engineering assistant. Deconstructs complex prompts into structured variables, generates bilingual vocabulary banks, identifies domain, evaluates quality, learns new terms, compares prompts, and optionally outputs PromptFill-compatible JSON. Use when a user wants to analyze, split, translate, learn from, compare, or template-ify any AI image/text prompt. Trigger phrases include: "analyze prompt", "split prompt", "extract variables", "create template", "learn vocabulary", "compare prompts", "词库", "提示词拆分", "提示词分析", "创建模板", "学习新词", "对比提示词".
srinitude/hermes-goal-prompt-generator/skills/goal-prompt-generator Skill
0★ repoGenerate isolated optimized Markdown goal prompts plus a paired ANALYSIS/BOOTSTRAP/RED/GREEN/REFACTOR TDD task list YAML that any coding agent can dynamically update through its building process. Use when a user asks to enhance, structure, validate, or save a goal prompt without executing it. Do NOT use as an automatic `/goal` preflight, slash-command interceptor, goal-loop hook, or to execute the generated goal or its task list; this skill only produces and validates the standalone prompt file and its companion task list YAML.
iTofu/handoff-skills/handoff-resume/skills/handoff-resume Skill
0★ repoUse when user wants to resume work from a previously saved handoff document, typically after running /compact, or when starting a session that continues earlier work. Triggers on "/handoff-resume", "从 handoff 恢复", "继续上次的工作", "resume handoff", "compact 之后继续".
Vladick-Pick/managing-multi-agent-teams Skill
0★Use when designing, auditing, or operating multi-agent teams, subagent delegation, supervisor/worker workflows, handoffs, agent crews, A2A/MCP-connected agents, long-running agent organizations, or coordination protocols.
rarexlabs/p11-public Skill
0★Create shareable, commentable p11 documents from agent output, then use reviewer comments to reply and revise. Use when an AI agent needs to hand plans, specs, proposals, or drafts to humans for review.
yystudio-cyber/agent-memory-blueprint Skill
0★Design, scaffold, validate, analyze, package, and publish machine-first memory blueprints for long-running AI agents. Use when Codex needs to create or improve a canonical multi-project memory YAML file, split execute/history/update read modes, define project skill cards and detail sections, maintain append-only event ledgers, validate strict-public safety, or prepare a public GitHub Agent Skill/template without leaking private content.
baslie/aitunnel-docs Skill
no license0★Локальная база знаний по сервису AITUNNEL (api.aitunnel.ru, docs.aitunnel.ru) — единый OpenAI-совместимый API к 100+ моделям ИИ (OpenAI, Anthropic Claude, Google Gemini, DeepSeek, Meta, Qwen) для пользователей из России без VPN. 45 страниц официальной документации: API, аутентификация, параметры, стриминг, лимиты, ошибки, prompt caching, structured outputs, tool calling, генерация изображений/видео/аудио, TTS, embeddings, rerank, reasoning, web search, PII, Responses API, Claude Messages API, OpenRouter из РФ, оплата, ключи, FAQ. Плюс 9 гайдов интеграции: Python OpenAI SDK, Cursor, Claude Code, Codex CLI, Qwen Code, OpenCode, OpenClaw, LangChain, Xcode. Используй когда упомянуты «aitunnel», «AITUNNEL», «aitunnel.ru», «sk-aitunnel-», вопросы про модели/цены/лимиты/ошибки aitunnel, «как подключить X к aitunnel», «base_url aitunnel», «OpenRouter в России», «Claude/GPT/Gemini в России без VPN». Также при «обнови документацию aitunnel», «refresh aitunnel docs», «пересобери docs» — для запуска scripts/update.ps1.