Skills
101,323 packaged capabilities and playbooks for agents. Install any of them with one command.
All
Hardware-offloaded erasure coding with DOCA Erasure Coding (Reed-Solomon, doca_ec) on a BlueField DPU, ConnectX NIC, or host: generate parity for data blocks, rebuild lost blocks from survivors, and refresh parity after a change. Trigger on “one data block changed, how do I refresh parity without re-encoding,” “a disk failed and 2 parity blocks are gone, can I rebuild,” “RAID-6 resilience across 12 disks,” “my doca_ec_task_create returns NOT_PERMITTED,” or “is this N+K layout still recoverable,” even without naming DOCA Erasure Coding. Covers bringing up a doca_ec context; create / recover / update tasks; matrix type / N / K / block size; doca_ec_cap_* queries before sizing; doca_mmap src/dst permissions; and DOCA_ERROR_* from doca_ec_task_*. Refuse and route elsewhere for non-Reed-Solomon codes (fountain / LDPC / raptor), pure-replication designs, network FEC, or other DOCA accelerator libraries (SHA / Compress / AES-GCM / DMA)
Use this skill for hands-on DOCA Ethernet packet-queue work on a BlueField DPU or ConnectX NIC—bringing up a `doca_eth_rxq` or `doca_eth_txq` on a port / representor / SF; choosing among the four `enum doca_eth_rxq_type` values (`_REGULAR` / `_CYCLIC` / `_MANAGED_MEMPOOL` / `_SHARED_MEMPOOL`); sizing burst or scatter-gather length against the `_cap_*` queries; submitting `doca_eth_txq_task_send` / `_lso_send` (carrying packet `doca_buf`s—no `doca_eth_frame` struct exists); or debugging DOCA_ERROR_* from an Ethernet call. `doca_eth_rxq` and `doca_eth_txq` are two Core contexts; one `doca_ctx_start()` does not cover both. Trigger on implicit phrasing: “my RX queue is up but no packets arrive,” “send-task returns AGAIN at line rate,” “which queue type for fixed-MTU ingress,” “device open fails without sudo,” or “is L3 checksum offload available here.” Refuse and route elsewhere for installing DOCA, flow-rule / steering programming, host↔DPU control messaging, or RDMA data movement
Audits any developer documentation site across 33 checks in 7 categories and produces a scored report (out of 100) with Pass / Warn / Fail status per check. Use this skill whenever a user provides a docs URL and asks to audit it, review it, score it, check its quality, or evaluate it for AI discoverability, SEO, structure, content, or completeness. Also trigger when the user says "run a docs audit", "audit these docs", "check this documentation", "how good are the docs for X", or pastes a docs URL with any evaluative intent. Always use this skill — do not attempt a freeform docs review without following this structured workflow.
Sắp xếp gọn kho tài liệu LOCAL (llmwiki/html/ + wiki/sources/draft/) khi phình to. Ba tầng — PROMOTE bản chất quý lên wiki ADR/concept (travel, sống sót clone), ARCHIVE render ephemeral (cặp proposal đã xong, bản bị thay thế, report cũ → archive/), KEEP canonical — rồi RE-INDEX đầy đủ. Trigger khi "dọn docs", "sắp xếp/merge tài liệu", "tidy docs", "html/draft quá nhiều", "/docs-curate".
Use this skill when users need to make early hires, build their founding team, determine compensation/equity, decide who to hire first, or scale from founders to first employees. Activates for "who should I hire first," "early hiring," "equity for employees," or team building questions.
AWS EC2 virtual machine management — instances, security groups, key pairs, AMIs, EBS volumes, Auto Scaling Groups, Spot Instances, Session Manager, placement groups, and instance lifecycle automation. Trigger on ANY of these, even when EC2 isn't named explicitly: - Launching or provisioning: "spin up a server", "create a VM", "new instance", "run-instances", mention of instance types (t3, m5, c5, r6, g5, p4d, t4g, c7g, etc.) - SSH / connectivity problems: "connection refused", "connection timed out", "permission denied publickey", "can't connect to my instance", "SSH not working" - Instance management: resize, stop, start, terminate, reboot, change instance type - Cost optimization: stop dev instances overnight, save money on EC2, spot vs on-demand, reserved instances - Auto Scaling: ASG, launch template, mixed instances policy, scale to zero, scheduled scaling - Spot Instances: spot fleet, spot interruption, capacity-optimized, price-capacity-optimized - AMIs and backups: create image, custom AMI, EBS snapshot, DLM lifecycle policy, copy AMI - Monitoring: EC2 CPU utilization, CloudWatch metrics for instance, instance status checks, console output - Access methods: Session Manager, EC2 Instance Connect, bastion host, port forwarding - Security: IMDSv2, instance metadata, IAM role on instance, security group rules - User data and bootstrap scripts, cloud-init
Professional UI design guidelines for accessible, well-structured interfaces. Use when: writing CSS, HTML, or frontend component code; designing websites, apps, dashboards, or any web interface; reviewing or improving existing UI designs; making decisions about colour, typography, layout, spacing, buttons, or forms; ensuring WCAG 2.1 AA accessibility; creating design systems or component libraries. Covers OKLCH colour palettes, 8pt spacing grid, fluid typography with clamp(), container queries, subgrid, form validation patterns, button hierarchy, dark mode, reduced motion, SEO meta tags, and Core Web Vitals.
Institutional-grade equity research report generation skill. Covers listed companies on China A-shares, Hong Kong stocks, and US stocks. Output modes: 1) Investment snapshot: 3-5 pages, including company overview, core financial metrics, valuation multiples, investment thesis, and risk factors; 2) In-depth research report: ≥25 pages, including industry analysis, industry chain mapping, three-statement model, DCF valuation, scenario analysis, and sensitivity testing. Trigger conditions: 1) Direct requests: analyze/look at/research/investigate/introduce + company name or stock; 2) Investment inquiries: what do you think/can I buy/is it worth investing/help me look/dig into + company or stock; 3) Keywords: research report, investment brief, investment snapshot, company one-pager, in-depth research report, in-depth research, in-depth analysis, individual stock analysis, one-pager, snapshot, company snapshot, equity research report; 4) Stock tickers: standard formats (e.g. 600519.SH, 0700.HK, AAPL, TSLA, BYD). When the intent is unclear, proactively ask for the output mode; never infer by default.
Use this skill when users need to build their positioning, develop their attractive character, or craft their origin story. Activates for "build my positioning," "create my origin story," "develop my brand," or when establishing expert authority.
Content-adaptive wisdom extraction that reads content first, detects which wisdom domains are present, and builds custom sections around them, with five depth levels and mandatory contrarian takes; pulls YouTube via fabric and articles via WebFetch. USE WHEN extract wisdom, analyze video, analyze podcast, extract insights, key takeaways, summarize interview, distill content. NOT FOR static Fabric extract_wisdom pattern (use Fabric).
Discover and import design tokens from SUBSCRIBED Figma team libraries into the current file. Use when the user wants to reuse tokens that live in a shared library: triggers 'list library variables', 'what tokens are in our shared library', 'import a variable from our design system library', 'pull in the brand color token from the library'. Works on ANY plan via the Plugin API teamLibrary methods. For LOCAL variables use figma-export-tokens / figma-manage-variables.
List a Figma file's version history, snapshot the file at any past version, and diff two versions (added/removed/renamed pages plus deep per-component changes). Use when the user wants to inspect Figma history — triggers: 'list Figma versions', 'what versions does this file have', 'show version history', 'snapshot this file at version X', 'get the file as it was last week', 'diff two Figma versions', 'what changed between these versions', 'compare version A and B', 'did this component change between releases'. Uses the Figma REST API + a personal access token because version history is NOT reachable via the Plugin API / use_figma / the native Figma MCP. For a prose release-notes writeup use figma-generate-changelog; to find which version introduced a change use figma-blame-node.
Visual quality review for FrameVideo compositions. Use when checking or fixing material overlap, text overlap, clipped or overflowing text, low text/background contrast, unsafe title/logo/CTA/caption placement, dynamic text fitting, visual collisions during animation, or final composition QA after creating or significantly editing FrameVideo HTML.
Use this skill when building applications with Gemini API hosted models, including Gemini and Gemma 4, working with multimodal content (text, images, audio, video), implementing function calling, using structured outputs, or needing current model specifications.
Access Gmail via CLI with 1Password OAuth: read, search, and export email, and create drafts. Replies quote the sender verbatim as styled blockquotes and can carry file attachments, built and verified in one command. TRIGGERS - gmail, read email, search inbox, export emails, create draft, compose email, reply to email, draft a reply, quote their words, quoted reply, attach a file to a draft.
Use when the user wants to learn Godot while building — teaching mode that explains the concept, the editor setup, and what to verify, instead of just delivering code. Triggers on "teach me", "explain as we go", "I'm learning Godot", "guide me", "walk me through", "help me understand".
Generate, reference, and conversationally edit short videos with Google's Gemini Omni Flash through the Gemini Developer API or Gemini Enterprise Agent Platform. Use when a task specifically needs Gemini Omni video, multimodal reference roles, time-directed prompting, audio-aware generation, Interactions API state, or a precise comparison with Veo, Nano Banana, consumer Gemini, or gateway routes.
Query and mutate content on the Hashnode GraphQL API (posts, drafts, publications, users, tags, series, comments). Use when a user wants to read or write data through the Hashnode gql endpoint, needs a reference for a specific query or mutation, is wiring up a Personal Access Token, or hits a "Publication does not have an active Pro plan" / FORBIDDEN error and needs to know an operation is Pro-gated.
ICP Filing (备案) rule consultation skill for Huawei Cloud. Based on Python 3.9+ with requests/Playwright, answers whether ICP filing is required, filing process & materials, account & entity limits, change filing, access filing, cancellation filing, APP filing, and migration/transfer filing. Read-only consultation, no filing operations on behalf of users. Use only when the user explicitly mentions 备案/ICP备案/备案规则/备案流程/接入备案/变更备案/注销备案/APP备案/迁移备案; refuses domain registration, DNS configuration, server purchase, real-name authentication, pricing. Trigger: 备案, ICP备案, 备案规则, 备案流程, 接入备案, 变更备案, 注销备案, APP备案, 迁移备案, 转移备案, 管局要求, 备案材料, 备案授权码
Remove signs of AI-generated writing from text. Use when editing or reviewing text to make it sound more natural and human-written. Based on Wikipedia's comprehensive "Signs of AI writing" guide. Detects and fixes patterns including: inflated symbolism, promotional language, superficial -ing analyses, vague attributions, em dash overuse, rule of three, AI vocabulary words, passive voice, negative parallelisms, filler phrases, and performed rigor. Includes a dedicated section for Slovak and Czech text, where most AI writing is a translated English draft and needs different punctuation, vocabulary, and word order rules.
End-to-end analytics & conversion tracking on Shopify Hydrogen — GTM, GA4 (browser + Measurement Protocol), Meta Pixel + CAPI, Google Ads, consent mode, CSP, Oxygen full-page cache. Real-world patterns from production deployments.
Owns the iteration loop on top of an ML workspace: the `journal/JOURNAL.md` index and the per-experiment `journal/NN_short_name.md` design notes that must be drafted and approved by the user **before** `experiments/NN_short_name.py` is created. Drives the propose → iterate → approve → implement → record loop; dispatches to `iterate-from-skore` / `iterate-from-user` for sourcing. TRIGGER — any of: - A session opens in an ML workspace (whether or not `journal/` exists yet — missing/placeholder → bootstrap mode). - User says "what's next", "resume", "where were we", "let's iterate", "propose next", "first baseline". - About to create a new `experiments/NN_*.py` (the matching `journal/NN_*.md` must exist and be approved first). - User wants to record an outcome from a finished run. - User asks to compare past experiments or review what's been tried ("compare X and Y", "where are we?"). SKIP when: no `journal/` yet AND no workspace scaffold (route to `organize-ml-workspace`); the work is mechanical inside `pipeline.py` / `evaluate.py` / `data.py` with no journal-level implication (owned by `build-ml-pipeline` / `evaluate-ml-pipeline`); the user asks for a symbol lookup (`python-api`); the user is diagnosing a single skore report without a "what next" framing (`evaluate-ml-pipeline`). HOW TO USE: read `journal/JOURNAL.md` first, classify the turn via the **Mode picker** (table near the top), then read only the matching section. Sibling skills open *just-in-time* when a step requires them — do not pre-read all sibling skills at session start. Design notes are the only artifact this skill writes; read, compare, and overview modes don't write.
Create and edit KubeSense dashboards over metrics, logs, traces, real user monitoring (RUM) and infrastructure host maps — directly with the create-dashboard and update-dashboard MCP tools, or as preset JSON the user imports — with the exact schema, the fields that hard-fail import, and the fields that silently discard your data instead of erroring. Includes validate-dashboard-json for checking a preset before you commit to it.
Generate high-converting landing pages as production-ready HTML with built-in design themes, copy frameworks (PAS, AIDA, BAB), CTA strategy, SEO meta, and automated conversion/speed audits. Use whenever the user wants a landing page, campaign page, lead capture page, promo page, launch page, or single-page site — or wants an existing landing page reviewed, audited, or improved — even if they don't say "landing page" (e.g. "a page for my product", "somewhere to send ad traffic", "a signup page for the webinar").
Guides engineering managers through the specific challenges of managing top engineers — produces a four-quadrant ability/confidence diagnostic, the Rock Star vs. Superstar distinction, common mistakes to avoid, a stagnation diagnostic (Diminishing XP), and a Pusher vs. Puller framework for managing burnout and team friction. Use when the user says "rockstar engineer," "superstar," "high performer," "brilliant jerk," "wants promotion," "hardest to manage," "overconfident," "my best developer is burning out," "engineer is frustrated," or "my best developer is pushing me." Do NOT use for standard underperformance (use performance-reviews) or general motivation questions (use engineer-motivation).
Helps EMs assess their own effectiveness, avoid common traps, navigate bad days, and handle recurring tensions in their own mindset and behavior. Use when the user says "I feel stuck," "am I doing this right," "personal development," "EM effectiveness," "blind spots," "bad days," "sanity check on my behavior," "the same problem keeps coming back," "made a mistake with someone," or "my team isn't motivated." Do NOT use when the issue is about the user's relationship with their own manager (use managing-up) or giving specific feedback to someone (use feedback).
When the user needs marketing ideas, inspiration, or strategies for their SaaS or software product. Also use when the user asks for 'marketing ideas,' 'growth ideas,' 'how to market,' 'marketing strategies,' 'marketing tactics,' 'ways to promote,' or 'ideas to grow.' This skill provides 139 proven marketing approaches organized by category.
Produce 3D assets with Meshy (meshy.ai) through its REST API and web platform — text-to-3D, image-to-3D, multi-image-to-3D, AI texturing/PBR, remesh/topology control, UV, and auto-rigging/animation. Use this skill when an agent must generate a mesh from a prompt or reference image, control polygon count and topology, apply or restyle textures, rig a humanoid for animation, export to GLB/FBX/OBJ/USDZ/STL for Unity/Unreal/Blender, drive Meshy's asynchronous job API (auth, polling, streaming, webhooks, credits, rate limits), review generated meshes for game/film/print readiness, or decide whether Meshy is the right tool at all. Not for editing an existing hand-modeled asset's geometry, hard-surface CAD with exact dimensions, or guaranteed-watertight manufacturing parts.
Use MiniMax's first-party Hailuo video API safely and reproducibly across global and mainland-China platforms. Covers current text-to-video, image-to-video, first/last-frame, and subject-reference models; prompt and camera craft; region/account routing; exact cost approval; asynchronous jobs and callbacks; artifact provenance; and rights, likeness, disclosure, and data-governance review. Use when designing, implementing, debugging, or evaluating direct MiniMax/Hailuo video-generation API workflows. Do not use for the consumer Hailuo site, Video Agent templates, gateways, or unrelated MiniMax modalities.
Use when a converted project (build_project --authored) needs its motion authored — turning motion.manifest.json + inert skeletons into real per-section/page motion with parity proof. Covers topology choice (one master scrub vs independent triggers), intent mapping (loader, journey, cursor, canvas, audio, ruler), per-target surfaces (Next hooks/globals vs Vite/Astro modules), reduced-motion, and the verification gate.
A UI motion reference library that helps you add suitable animations to your product. No need to understand motion design — describe your product scenario (slow page load, no response after a user clicks, wanting the page to feel more polished) and get a recommended approach plus code. Trigger phrases: 页面没动画、交互没反馈、怎么加动效、让页面动起来、加载太慢怎么办、想让产品高级一点、motion ref.
Use when a topic or news stream needs recurring monitoring — cover it with web search, multi-language sources, digest formatting, and automated delivery via Hermes cron jobs. Covers search strategy, source selection, Chinese-language platforms, and digest templates.
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.
