Skills
97,501 packaged capabilities and playbooks for agents. Install any of them with one command.
All
Conducts an independent review of a result specification or implementation plan without fixing them. Use when the user asks to check a spec or plan, a CreatePlan card, acceptance criteria, or the scope of a task. Do not use when the user asks to write a specification or plan, implement a task, fix an artifact, or review code changes.
Review commercial proposals and build customer-led business cases, ROI or value models, pricing or investment rationales, executive summaries, and customer-ready value stories tied to a customer, workflow, initiative, or decision.
Builds robust Python MCP (Model Context Protocol) servers with FastMCP—covering tool design, error contracts, event-loop-safe blocking work, subprocess/CLI wrapping, single-file vs. packaged distribution, global-state-free testing, and prompt-injection awareness. Use when writing an MCP server, exposing a tool or CLI to an LLM client, debugging tool registration or packaging, or testing MCP tools.
Use PROACTIVELY when users ask about React project structure, Bulletproof React patterns, or need architecture guidance. Covers structure setup, codebase auditing, anti-pattern detection, and feature-based migration planning. Triggers on "bulletproof react", "React structure help", "organize React app", or "audit my architecture".
Headless browser testing via persistent Bun.WebView session. Use when testing frontend UI, verifying page content, filling forms, clicking buttons, taking screenshots, or interacting with local or staging web applications. Triggers on phrases like "test the page", "check the UI", "browse to", "fill the form", "click the button", "take a screenshot".
Implement the Business Brain pattern. Gives every agent skill access to your tone, audience, and positioning without bloating the context window. Use when asked for brand context, voice, or positioning, or when setting up a new project's brain.
Researches software/SaaS products, business reputation, and crowdfunding campaigns via the Crawlora API: Product Hunt launches/makers/alternatives, Trustpilot business reviews, TrustMRR revenue-verified startups, Capterra software reviews, BBB (Better Business Bureau) business profiles/complaint history/Scam Tracker reports, and Kickstarter campaign discovery/funding/updates/comments—returning clean JSON. Use when the user wants a product's launch/review history, a company's Trustpilot or BBB reputation, a startup's verified revenue, software alternatives/reviews before buying, or a crowdfunding campaign's funding status and backer updates.
Volcano Engine AI MediaKit talking-head video editing Skill: a one-stop workflow from environment setup through media management, audio processing, talking-head cuts, video export, review UI, and iterative refinement. You MUST invoke this Skill when the user mentions talking-head editing, cutting talking video, video editing, removing pauses, processing audio, exporting talking video, automatic editing, removing verbal slips, or similar. Also invoke when the user uploads video or audio and asks for editing.
Produce a capacity planning document for a service covering traffic forecasts, resource requirements, and scaling strategy. Use when asked to plan infrastructure capacity, forecast resource needs, model traffic growth, define scaling strategy, or produce a capacity review for a service. Produces a structured capacity plan covering current baseline metrics, growth projections, resource requirements per tier, scaling strategy, cost projections, capacity triggers, and an infrastructure action roadmap.
An ongoing planning skill for personal career and professional development. Use when the user discusses work uncertainty, strengths, interests, long-term direction, role or industry choices, promotion, career change, leaving or starting a job, offers, job descriptions, entrepreneurship, freelancing, layoffs, organizational change, industry decline, or AI replacement risks; they need not explicitly ask for career planning. The core task is to extract proven abilities from real experiences and outcomes, identify work they are willing to repeat long term, map it to specific roles and industries, and provide a primary path, adjacent alternatives, career moat, master’s lens, antifragility and ABZ, and an action roadmap. At the start of a full consultation, if unspecified, first ask whether they have Feishu CLI. Once CLI availability and authorization are confirmed, let them choose between local HTML and a Feishu document with a main canvas. If information is insufficient, ask 1–3 key diagnostic questions, then generate or update the selected career profile after the user answers. Not for requests limited to writing weekly reports, polishing copy, performing technical tasks, or asking general management questions, unless they affect the user’s personal career choices or long-term skill development.
Computes the best-response allocation of roster resources across categories in a Head-to-Head Categories matchup. Given our per-category capacity, the opponent's projected output, per-category win probabilities (from matchup-win-probability-sim), and a K-of-N winning threshold, classifies categories into pushed / contested / conceded buckets, emits per-category leverage weights for downstream lineup and streaming decisions, computes the resulting K-of-N win probability, and writes a plain-English rationale. Domain-neutral — portable to any fantasy sport with H2H Cats scoring (MLB 10-cat, NBA 9-cat, NHL 10-cat). Use when you need push/punt decisions, dominated-strategy elimination, leverage weights per cat, or best-response allocation; or when the user mentions "category allocation", "push or punt", "K of N cats", "dominated strategy elimination", "best response allocation", "Blotto fantasy", "leverage weights per cat", or "which cats to push".
Manages California Consumer Privacy Act (CCPA) consumer rights requests under Civil Code sections 1798.100-125, covering the right to know, right to delete, right to opt-out of sale, and non-discrimination. Includes 45-day response window and identity verification requirements. Activate for CCPA, California privacy, right to know, right to delete, opt-out of sale queries.
This skill provides Fountain format character and dialogue formatting rules. Covers character name formatting, extensions (V.O., O.S., CONT'D), dual dialogue, parentheticals, and proper dialogue block structure. Use when: introducing characters, writing dialogue blocks, formatting character names, handling dual dialogue, or using character extensions.
When writing reports in Feishu Docs, create, edit, and deliver chart-assistant editable chart cards. Covers 30+ built-in chart types, including columns/bars (stacked, grouped, percentage, and overlapping comparisons), line, area, scatter/bubble, dual-axis, pie/doughnut, rose, radar, funnel, word cloud, Sankey, heatmap, Treemap/Sunburst, as well as business charts such as waterfall/bridge, Mekko, bullet, dumbbell, butterfly, and small-multiple panels. Chart types, annotations, and editing capabilities are broadly aligned with ThinkCell. By default, follows MBB (McKinsey/BCG/Bain) consulting-style charting standards and provides report-grade analytical annotations such as variance annotations (total variance, hierarchical variance, compound growth rate), trendlines, reference lines/regions, and highlight coloring. Supports connecting Feishu Sheet/Base/Fengshen data sources and keeping them in sync, recreating charts from reference images, creating Feishu Docx files, and inserting/replacing chart-assistant cards in existing documents; it can also explain and validate chart-assistant configurations. Use when the user explicitly mentions Chart Assistant, editable charts or configurations, built-in Chart Assistant templates, Chart Assistant cards, data-source sync, turning Fengshen or Feishu Sheet/Base links into editable charts, or requests MBB/consulting-style charts or recreation from reference images, even if they don't name the skill. Do not take over ordinary requests like "make this a chart" or "draw a trend chart" unless they require the product capabilities above; if the user only wants a text explanation and no chart, don't trigger the chart-creation flow.
Exit ritual for a work session. Use when the user invokes /checkpoint, says they are done for the day or leaving a project for a while, or accepts the wrap-up proposal armed by /lets-work. Routes each piece of understanding gained this session to its proper home — a code comment, the PR/commit message, or the personal journal — then records the visit marker the next /lets-work will brief from.
Run chief-agent across as many batches as it takes to finish a milestone, writing one report per task instead of one per batch. When a task hits ambiguity, a throwaway decision-support agent proposes options; chief-agent still makes the final call and the report captures the reasoning. Requires goals and contracts to exist. Use "/chief-loop".
A complete toolkit for generating Chinese software copyright application materials. Supports automatically extracting information from project code and documentation to generate software copyright registration application forms, source code documents (30 pages each from the front and back), user manuals and design specifications, and automatically converting them to PDF files. Suitable for WeChat Mini Programs, web apps, mobile apps, desktop apps and other software projects. Use this skill when the user needs to apply for a Chinese software copyright.
Daily autonomous classifier for failing PRs across your repos. Runs /ci-debug headless against every open PR with red required checks, posts the verdict as a collapsed PR comment, and appends to a per-repo .sentinel/ledger.jsonl. v1 is propose-don't-apply — NEVER auto-pushes a fix. Use when you're tired of /status sweeps catching the same 10 CI failure patterns over and over.
Query California Irrigation Management Information System (CIMIS) weather data using the cimis CLI. Use when working with irrigation data, weather stations, ET0 (evapotranspiration), or when the user mentions CIMIS, California weather, irrigation scheduling, or cimis.
Verify that every citation in a manuscript is real, correctly attributed, and accurately described. Detects ghost papers, wrong arXiv IDs, inverted claims, and dead links by fetching each cited work. Optional fix mode applies bib metadata corrections and surfaces prose rewrites for claim errors. Triggers on: "check my citations", "verify references", "citation audit", "are my references real", "check bib", "reference check", "bib audit", "citation verification". Companion to manuscript-review (Pass 5 hygiene); this skill audits factual truth.
Comprehensive citation management for academic research. Search Google Scholar and PubMed for papers, extract accurate metadata, validate citations, and generate properly formatted BibTeX entries. This skill should be used when you need to find papers, verify citation information, convert DOIs to BibTeX, or ensure reference accuracy in scientific writing.
(NS) Resolve ambiguities, gaps, and unstated assumptions in a scope description before generating requirements. On brownfield repos, runs bootstrap-brownfield when brownfield-map.md is missing; when a map exists, mandatory human gate to refresh or keep it (shows last map date). Use when the user gives a vague or incomplete scope, mentions "dashboard", "notifications", "integration with X" without detail, or before requirements generation when terms are ambiguous — even if they do not ask to "clarify". Do NOT generate requirements.md in this skill; output enriched context only.
Activate a 4-stage coding discipline framework that forces Claude to plan before coding, isolate changes on a branch, write tests first, and self-review output twice before presenting it. Use when starting a complex coding task, when past Claude sessions produced broken first drafts, or when you want to prevent rework cycles. Produces a confirmed written plan, isolated feature branch, test-first implementation, and a double-reviewed output with a correctness and code-quality checklist.
Audio processing for video files using FFmpeg. Loudness normalization to streaming standards (LUFS), noise reduction, audio mixing, extraction, replacement, silence detection and removal, equalization, compression, fade effects, format conversion, and audio ducking. Use when user says "audio", "normalize", "volume", "loudness", "LUFS", "noise", "silence", "extract audio", "replace audio", "mix audio", "equalizer", "compress audio", "ducking", or "sidechain".
Only use when the user explicitly invokes /cleanup-storage (or $cleanup-storage in Codex). Never auto-invoke. A skill for relieving storage pressure on a local machine. Examines disk usage, presents deletion candidates grouped by safety category, deletes only after getting the user's approval for each category, and reports the change in available space.
Use when a val needs to ship JavaScript that runs in the browser — React apps, vanilla DOM scripts, canvas/games, htmx/Alpine, or any client-side module beyond a single inline snippet. Explains how Val Town serves transpiled .ts/.tsx/.jsx modules with no build step, how the browser resolves their imports, and how to load third-party deps.
Prepare and validate research-only clinical decision-support evaluation, evidence-profile, cohort, survival, biomarker/model, privacy, and governance artifacts. Use for aggregate or synthetic research documentation and traceability—not patient care or live clinical operation.
Guide to the Cloudflare Agents SDK for building stateful AI agents on Workers plus Durable Objects. Covers the Agent class, this.setState and this.sql state, WebSocket and streaming chat via AIChatAgent and useAgentChat, scheduling with this.schedule (delays, dates, cron), triggering Workflows, Browser Rendering, RAG with Vectorize and Workers AI embeddings, MCP servers with McpAgent, human-in-the-loop review, and routing through routeAgentRequest, getAgentByName, useAgent and AgentClient. Documents 15+ pitfalls: missing new_sqlite_classes, non-atomic migrations, Agent class not exported, binding name mismatches, global uniqueness surprises, state size limits, MCP transport confusion and bypassed authentication. Use when building AI agents, chat agents, MCP servers or multi-agent systems on Cloudflare, or debugging Agent binding, migration and state-sync errors.
Manage Cloudflare DNS records and zones from the terminal with the cf CLI (coollabsio/cloudflare-cli). Full DNS CRUD, zone listing, filtered queries, proxied record toggling, and JSON output for scripting. Use when: managing DNS records via CLI, listing Cloudflare zones, creating/updating/deleting DNS records, automating Cloudflare DNS workflows, or troubleshooting zone-scoped token permission errors, “cf: command not found,” and config file issues
Complete knowledge domain for Cloudflare D1 - serverless SQLite database on Cloudflare's edge network. Use when: creating D1 databases, writing SQL migrations, configuring D1 bindings, querying D1 from Workers, handling SQLite data, building relational data models, or encountering "D1_ERROR", "statement too long", "too many requests queued", migration failures, or query performance issues. Keywords: d1, d1 database, cloudflare d1, wrangler d1, d1 migrations, d1 bindings, sqlite workers, serverless database, edge database, d1 queries, sql cloudflare, prepared statements, batch queries, d1 api, wrangler migrations, D1_ERROR, D1_EXEC_ERROR, statement too long, database bindings, sqlite cloudflare, sql workers api, d1 indexes, query optimization, d1 schema
Guide to Cloudflare Durable Objects: globally unique, stateful objects for coordination, real-time communication and persistent state. Covers the DurableObject class and bindings, WebSocket Hibernation (ctx.acceptWebSocket, webSocketMessage, serializeAttachment), SQLite storage via ctx.storage.sql, alarms with storage.setAlarm and the alarm() handler, ids from idFromName, newUniqueId and getByName, RPC methods, blockConcurrencyWhile, migrations and location hints. Documents 15+ pitfalls: class not exported, missing or wrong migrations, new_sqlite_classes, constructor work and setTimeout or fetch breaking hibernation, in-memory state lost on hibernation, partial deleteAll on the KV backend, binding mismatches, state size limits and alarm retry failures. Use when building chat rooms, multiplayer games, rate limiters, session stores, queues or per-user state on Workers, or debugging DO migration, hibernation and alarm errors.
Comprehensive Cloudflare account management for deploying Workers, KV Storage, R2, Pages, DNS, and Routes. Use when deploying cloudflare services, managing worker containers, configuring KV/R2 storage, or setting up DNS/routing. Requires CLOUDFLARE_API_KEY in .env and Bun runtime with dependencies installed.
Use this skill when a product engineer wants to build, migrate, or extend UI using the Earth Design System (EDS, `@adaptavant/eds-core`). Triggers include any mention of 'build with EDS', 'use EDS', 'EDS component', 'migrate to EDS', 'convert to EDS', references to `@adaptavant/eds-core`/`@adaptavant/eds-icons`, or requests like "add a form/page/modal using EDS". Acts as a co-author/buddy that builds working EDS-compliant JSX immediately from a clear request, enforcing EDS guardrails along the way. Do NOT use this skill for fixing bugs inside the EDS library itself (use `/bug-fix-workflow`) or for writing PR descriptions (use `pull-request-author`).
Use when receiving code review feedback (especially if unclear or technically questionable), when completing tasks or major features requiring review before proceeding, or before making any completion/success claims. Covers three practices - receiving feedback with technical rigor over performative agreement, requesting reviews via code-reviewer subagent, and verification gates requiring evidence before any status claims. Essential for subagent-driven development, pull requests, and preventing false completion claims.
Before pushing a non-trivial change, spawn read-only subagents on two different model families to adversarially review the same diff in parallel, then merge the findings yourself. Use for "multi-model review", "adversarial review", "challenge this", or a pre-push review of a large or risky diff — not for small or low-risk changes, which don't justify the cost of a panel. Once a PR exists, cloud bot feedback is handled by `pr-review-loop`, not this skill
Review one implementation change against its Phat task, feature spec, acceptance criteria, repository engineering standards, and any accepted prototype evidence. In standalone or final mode record supported task evidence/status; in delegated mode return a read-only report to the parent. Use after implement-task or when reviewing a working tree, staged change, commit, or branch.
Performs OWASP-based code security audits on any codebase. Analyzes source code against ASVS 5.0.0 verification requirements, API Security Top 10 2023 risk patterns, OWASP CheatSheet secure coding practices, and WSTG testing methodologies. Input is a codebase to review; output is a detailed Markdown security audit report. Use when the user requests a security audit, security review, vulnerability assessment, or code security analysis.
Inspect an existing repository and refresh only the generated section of CONTEXT.md with attributable Observed, Inferred, and Unknown evidence. Use for onboarding, material repository-context gaps, or explicit refresh; not as a mandatory stage during ongoing implementation.
MANDATORY handoff to the local Codex CLI. Triggers when (a) the user's message begins with `/codex` as a command, or (b) the user issues an explicit handoff directive like "hand this to codex", "run this through codex", "ask codex", or "have codex do X". On trigger, pipe the user's prompt verbatim to `codex exec` and return Codex's final message verbatim. Treat this like a shell alias the user is executing through you. Do NOT interpret the task, inspect files, gather context, attempt the work yourself, judge whether Codex is the right tool, or rewrite the prompt. The only abort condition is an empty prompt (ask what Codex should do).
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.