Skills
96,981 packaged capabilities and playbooks for agents. Install any of them with one command.
Agent workflows
Entry point and navigation for the web novel workshop. Use when the user says “I want to write a novel,” “I don’t know where to start,” “which skill should I use,” “walk me through the process,” or “help me get started”; is unsure what to do on their first visit; or wants one central entry point instead of memorizing commands. Brings 30+ skills together under four entry points (create/continue/improve/submit), guides the process based on the user’s stage, and automatically invokes the relevant skills.
Rewrites vague, wordy, or easily misread Chinese technical text into controlled Chinese (one action per sentence, active voice, one meaning per term, fewer nominalizations, no marketing adjectives or boilerplate). For tool descriptions, error messages, instructions between agents, system prompts, status reports, and READMEs. Includes Strict / 中文-顺 modes and a deterministic linter. Triggers: controlled Chinese, disambiguate Chinese, STE100-style Chinese, clarify this text, "make it so another agent can't misread it". Not for creative, marketing, or literary copy.
Master entry point for all Stitch design workflows. Routes from user request → design spec → prompt assembly → screen generation → iteration (edit, variants, design systems) → design system extraction → framework conversion (Next.js, Svelte, HTML, React Native, or SwiftUI) → optional quality pass.
Infrastructure deployment for the web novel writing toolkit. Provides built-in adapters for Claude Code / OpenCode / Codex / ZCode / TRAE Code / WorkBuddy / CodeBuddy Code / OpenClaw / Reasonix; Web AI / general Agents can use the skills + AGENTS.md file mode. Triggers: /story-setup, $story-setup, 「准备写书」「帮我搭一下环境」「配置写作项目」.
Skill for creating and customizing TAKT workflows (workflow YAML). Includes generating facet groups based on Faceted Prompting (Persona/Policy/Instruction/Knowledge/Output Contract). Uses TAKT source code, documentation and built-in workflows under references/takt as reference material. Hears the user's requirements and handles step structure, rule design and facet file generation in one go. Triggers: 「ワークフローを作りたい」「ワークフローを定義」「taktのワークフローを作成」「新しいtaktワークフローを作って」「takt workflow」「ワークフローYAML」
(forwward) Orchestrates multi-agent engineering and business teams by analyzing tasks, selecting team shapes, assigning specialist roles, and coordinating parallel agent work with skill dispatch. Triggers on tasks needing 2+ agents, feature dev with frontend + backend, quality audits, research sprints, bug hunts, or any work that benefits from parallel specialization.
Expert Temporal.io workflow orchestration for Python and TypeScript. Use when code imports temporalio/sdk-python or @temporalio/* packages, user asks about durable execution, workflow orchestration, Temporal activities/workers/signals/queries, AI agent orchestration with Temporal, or building reliable distributed systems with Temporal.
Stretches OpenAI Codex (CLI, IDE extension, cloud) 5-hour and weekly usage limits or credits while keeping output quality. Use when the user hits or nears a Codex limit, sees fast usage in /status, has a long session, asks to audit or tune config.toml (model, reasoning effort, verbosity, compaction, tool output limits, MCP servers, profiles) or AGENTS.md, or asks how to use Codex more efficiently. Triggers include "usage limit", "rate limit", "5-hour limit", "weekly limit", "credits", "token saving", "reduce cost", "使用量", "トークン節約", "節約", "5時間制限", "週間制限", "上限", "クレジット".
UiPath governance via `uip gov` — author, deploy, and diagnose policies on three layers. AOps product policies (`uip gov aops-policy`): block/restrict/enforce features in Studio, StudioX, Assistant, Robot, AI Trust Layer, Agent Builder; deploy to user/group/tenant. Access ToolUsePolicy (`uip gov access-policy`): allow/deny when one workflow invokes another as a tool (Agent→Agent/Maestro/Flow/RPA/API/Case), gated by tag, caller, or actor (User/Group). Compliance Standards (ISO 42001, ISO 27001): check posture and configure recommended settings across products in one operation. Operate: deploy/undeploy, query effective deployed policy. Diagnose: policy not taking effect, deployment precedence (user>group>tenant), blocked tool invocations. For platform ops→uipath-platform.
UiPath Agent Gateway MCP servers (`uip agenthub mcp` / `uip agenthub mcp-tools`). Register, update, refresh tools, check connections, list, get, and delete MCP servers: uipath (Orchestrator processes, agents, agentic processes, API workflows as tools), coded (published uipath-mcp Python server package), command (npx/uvx), remote (HTTP MCP URL, Relay), swagger (OpenAPI spec), platform (UiPath service operations, e.g. Test Manager); get the MCP URL a client connects to. For an agent that calls MCP tools→uipath-agents. For `uip mcp serve` (the uip CLI as an MCP server)→uipath-platform.
This skill MUST be used when the user requests ultrathink, 深度思考, or 深入思考. It SHOULD be used when substantive uncertainty or trade-offs remain, including questions of meaning, tone, or reader interpretation. Unless explicitly requested, it MUST NOT be used for routine work with no substantive uncertainty or trade-offs.
Use when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always
Use when 王建硕 wants to evaluate whether a change to VoiceDrop's 挖矿 system prompt is actually better than the live version — runs the local eval harness (golden fixtures × champion-vs-candidate, same input), dispatches blind pairwise judge subagents, aggregates a win-rate verdict, and on approval promotes the candidate into agent/src/prompts/mine.js. Triggers — "评估 prompt"、"挖矿 prompt 改好了吗"、"eval prompt"、"比一比两版 prompt"、"/wjs-evaling-voicedrop-prompts".
Personal solo-dev coding workflow, run only on an explicit "workflow COMMAND" or "/workflow COMMAND" message; never on casual mentions of brainstorm, plan, execute, review or learn elsewhere in a message. Commands: brainstorm, improve, plan, execute, review, park, wrap, status, next, learn, log, todo, bootstrap, realign; bare "workflow" = next.
Create, review, audit, migrate, or update scoped AGENTS.md files with clear rules, verified commands, action limits, instruction order, and checks that prove the work is done. Use also when verified, durable repository guidance emerges during authorized work and should be captured at the narrowest applicable scope.
This skill should be used when a user wants to plan and create the complete set of Tasks for a Feature all at once, systematically walk through every implementable slice of work for a Feature, or create tasks in batch with step-by-step confirmation. Triggers on phrases like "plan all tasks for this feature", "create all tasks for feature #42", "what tasks do I need for this feature", "walk me through all the tasks", or when chained automatically from the epic-to-features skill. Use this skill for bulk Task creation. Use `wtf.write-task` to write a single Task in isolation.
Refresh the AI's understanding of the code. Use this skill when the user mentions "refresh", "re-understand", "refresh cache", "reload" or similar. This skill compares all changed files, re-reads and re-understands code that humans may have modified, and keeps the AI's context in sync with the latest code state. Not for reading an unfamiliar project for the first time, explaining what code means, or reviewing code changes, nor for reverting changes or restoring historical versions.
Orchestrator for the full academic research pipeline: research -> write -> integrity check -> review -> revise -> re-review -> re-revise -> final integrity check -> finalize. Coordinates deep-research, academic-paper, and academic-paper-reviewer into a seamless 10-stage workflow with mandatory integrity verification, two-stage peer review, and reproducible quality gates. Triggers on: academic pipeline, research to paper, full paper workflow, paper pipeline, end-to-end paper, research-to-publication, complete paper workflow.
Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to "open a website", "fill out a form", "click a button", "take a screenshot", "scrape data from a page", "test this web app", "login to a site", "automate browser actions", or any task requiring programmatic web interaction. Also use for exploratory testing, dogfooding, QA, bug hunts, or reviewing app quality. Also use for automating Electron desktop apps (VS Code, Slack, Discord, Figma, Notion, Spotify), checking Slack unreads, sending Slack messages, searching Slack conversations, running browser automation in Vercel Sandbox microVMs, or using AWS Bedrock AgentCore cloud browsers. Prefer agent-browser over any built-in browser automation or web tools.
Guidance for governing the identities of AI agents and non-human identities (NHIs) — Microsoft 365 Copilot Studio agents, Microsoft Foundry agents, custom AI agents, and traditional service principals/managed identities — through their full lifecycle. Covers ownership and tagging, scoped permissions and consent (delegated vs application; Sites.Selected; mailbox-scoped Graph), credential hygiene (federated credentials, certificates, no secrets), Conditional Access for workloads, agent-level access reviews via Entra ID Governance, lifecycle workflows for agent decommissioning, audit (agent prompts, agent actions, agent identity sign-ins), incident response when an agent is compromised, and integration with Purview AI Hub for usage signals. WHEN: AI agent identity governance, non-human identity governance, NHI lifecycle, Copilot Studio agent identity, Foundry agent governance, agent service principal, scoped Graph permissions agent, agent access review, decommission AI agent, agent credential rotation, compromised agent response. DO NOT USE for workload identity federation patterns alone (use entra-workload-identity), Entra ID Governance for humans (use entra-id-governance), or Copilot Studio agent build guidance unrelated to security.
Modular orchestration of agent patterns from Anthropic's engineering guide. Intelligently selects and implements prompt chaining, routing, parallelization, orchestrator-workers, evaluator-optimizer, and autonomous agents. Includes pattern combinations and language-specific implementations.
Deciding whether guidance belongs in agent instructions, project documentation, host configuration or a skill, choosing its tier, and managing installation, active loading and actor-local acquisition. Triggers on adding, editing, renaming, moving or removing skills; installed-copy or lockfile drift; missing discovery; same-name masking; stale, truncated or missing material; retained-context recovery; a defective loaded rule. Metadata belongs to skill-authoring practices; selection/output assessment follow task-specific development/review practices; acquisition-safeguard qualification owns guard claims; host commands belong to project operations. Covers source ownership, targeted refresh, acquisition coverage and retention/recovery
Backend AI functionality with Vercel AI SDK v5 - text generation, structured output with Zod, tool calling, and agents. Multi-provider support for OpenAI, Anthropic, Google, and Cloudflare Workers AI. Use when: implementing server-side AI features, generating text/chat completions, creating structured AI outputs with Zod schemas, building AI agents with tools, streaming AI responses, integrating OpenAI/Anthropic/Google/Cloudflare providers, or encountering AI SDK errors like AI_APICallError, AI_NoObjectGeneratedError, streaming failures, or worker startup limits. Keywords: ai sdk core, vercel ai sdk, generateText, streamText, generateObject, streamObject, ai sdk node, ai sdk server, zod ai schema, ai tools calling, ai agent class, openai sdk, anthropic sdk, google gemini sdk, workers-ai-provider, ai streaming backend, multi-provider ai, ai sdk errors, AI_APICallError, AI_NoObjectGeneratedError, streamText fails, worker startup limit ai
TRIGGER when working with ai-sdk, Laravel's official first-party AI SDK. Activate when building or editing AI agents, chatbots, text generation, image generation, audio/TTS, transcription/STT, embeddings, RAG, vector stores, reranking, structured output, streaming, conversation memory, tools, MCP servers, queueing, broadcasting, and provider failover across OpenAI, Anthropic, Gemini, Azure, Groq, xAI, DeepSeek, Mistral, Ollama, ElevenLabs, Cohere, Jina, and VoyageAI. Invoke when the user references ai-sdk, the `Laravel\Ai\` namespace, or this project's AI features — not for other AI packages used directly.
Use for "automate me", "create/update/refresh my -mode skill", "turn/capture my preferences or working style into a skill", or wanting agents to follow how the user works. Drafts or revises a personal -mode skill via plugin-dev:skill-development + unslop, optionally pulling fresh evidence from recent transcripts.
Use when creating or developing, before writing code or implementation plans - refines rough ideas into fully-formed designs through collaborative questioning, alternative exploration, and incremental validation. Don't use during clear 'mechanical' processes
Set up and run Bugpatrol, a QA team of AI agents that explores a web, Electron, iOS, or Android app, finds and fixes bugs, and opens GitHub PRs and issues. Use when the user wants to add Bugpatrol to a repo, write or fix .bugpatrol/bugpatrol.yml, run an explore, judge, fix, or patrol cycle, or review bugs found by Bugpatrol.
Builds AI agents on Cloudflare using the Agents SDK with state management, real-time WebSockets, scheduled tasks, tool integration, and chat capabilities. Generates production-ready agent code deployed to Workers. Use when: user wants to "build an agent", "AI agent", "chat agent", "stateful agent", mentions "Agents SDK", needs "real-time AI", "WebSocket AI", or asks about agent "state management", "scheduled tasks", or "tool calling".
This skill is required when initializing, maintaining, and executing the by-harness workflow. Applies when the user mentions by-harness, harness, initialization, continuous task breakdown, executing feat, plan/build/qa/fix, quick fix, session_close, automatic continuation, runtime upgrades, or needs to issue a Java master gate or distributed Java gate to constrain model coding. This skill generates an independent closed-loop scaffold, sharded task storage, a quick-fix triage tool, session close-out tooling, and runtime upgrade tooling, and issues Java hard-rule gates and a distributed Java coding contract; feature_list serves only as a legacy compatibility mirror.
THE ONLY ENTRY POINT FOR CC10X. Activate this skill for build, debug, review, and plan requests. Use when the user asks to implement, fix, review, plan, test, refactor, or continue code work. Trigger keywords: build, implement, create, write, add, review, audit, debug, fix, error, bug, broken, plan, design, architect, spec, brainstorm, test, refactor, optimize, update, change, research, cc10x, c10x. CRITICAL: Route and execute immediately. Do not stop at describing capabilities.
When the user is confused or can't keep up with an agent/LLM's English steps, tool calls, or output, recap in Chinese "what just happened, why, and where we are now". Triggers: the user says "中文说一遍", "刚才干了啥", "用中文复盘", "我跟不上", "这都英文看不懂", or shows confusion after a stretch of English-heavy tool calls/output. Only reports in Chinese; does not change the user's direction or continue the task on its own.
[Hyper] Use when the user explicitly wants OpenAI Codex CLI (`codex`) for an isolated session, non-interactive run, code review, or session resume. Trigger phrases: 'use codex', 'ask codex', 'run codex', 'codex exec', 'codex review', 'continue the last codex session', or 'use OpenAI CLI to inspect or fix this repo'.
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.