Skills worth trusting.

kubernetes-operator

Use when building a Kubernetes Operator — custom controllers that reconcile CRD state. Triggers on "build an operator", "CRD design", "reconcile loop", "controller-runtime", "kubebuilder", "operator-sdk", "metacontroller", "KOPF", "operator capability levels", or "custom resource". Ships CRD validator, reconcile-loop linter, and OperatorHub capability auditor (all stdlib Python), 4 references on the operator pattern + CRD design + reconcile patterns + tooling landscape, and a /operator-audit slash command. NOT a generic k8s skill — specifically the Operator pattern.

better-logging

Design durable, queryable operation outcome events for HTTP, IPC, tRPC, jobs, startup flows, and queues so failures and latency are easy to investigate.

oiloil-ui-ux-guide

梳理项目 UI/UX 设计规范并输出 design-spec.md，评审现有界面并给出按优先级排序的修复建议，或整理指定界面的设计规则。用户要求设计系统、设计规范或完整界面评审时使用；单个颜色、字号或按钮大小的简单问题直接回答。

braintrust-tracing

Braintrust tracing for Claude Code - hook architecture, sub-agent correlation, debugging

verify

Use when validating implementation against spec artifacts before archive — not for design, planning, or implementation

bee-build-and-release

prd

Generate a Product Requirements Document (PRD) for a new feature. Use when planning a feature, starting a new project, or when asked to create a PRD. After PRD is confirmed, use /prd-to-spec (optional) for technical design, then /to-issues to create implementable tickets. Triggers on: create a prd, write prd for, plan this feature, requirements for, spec out, 写PRD, 需求文档, 需求分析, 规格说明.

avalonia-viewmodels-zafiro

Optimal ViewModel and Wizard creation patterns for Avalonia using Zafiro and ReactiveUI.

humanize-it

对指定文档进行去 AI 味的改写。自动选择最合适的人性化策略（humanizer-zh / humanize-chinese / technical-writing）， 迭代改写直到效果达标或迭代 42 次为止。适用于中文文本的去 AI 化处理，包括通用文章、技术文档、学术论文等。 Use when user says: "humanize this", "去AI味", "降AIGC", "人性化改写", "改成人话", "去除AI痕迹", "humanize document", "make text human-like", "去机器味", "降低AI率", "过AIGC检测"

longbridge-quant-stats

Quantitative statistics framework for time-series analysis using Longbridge price data — ADF unit root test (stationarity), cointegration (Engle-Granger / Johansen), GARCH volatility modelling (conditional heteroskedasticity), regression diagnostics (Durbin-Watson / Breusch-Pagan), bootstrap confidence intervals, hypothesis tests (t-test / F-test). Requires statsmodels and scipy. Triggers: "量化统计", "ADF检验", "单位根", "协整检验", "GARCH", "自相关", "异方差", "Bootstrap", "假设检验", "量化統計", "ADF檢驗", "單位根", "協整檢驗", "異方差", "假設檢驗", "quantitative statistics", "ADF test", "unit root", "cointegration", "GARCH", "autocorrelation", "heteroskedasticity", "bootstrap", "hypothesis test", "statsmodels".

llmquant-etfs

Router skill for LLMQuant ETFs workflows. Use when the user needs ETF holdings, overlap, concentration, issuer snapshot, or theme exposure analysis.

longbridge-seasonality

Seasonality and calendar-effect strategy via Longbridge Securities — uses historical OHLCV data to compute month-of-year returns (January Effect), day-of-week returns (Monday / Friday effect), pre/post-holiday drift, and earnings-season effect; identifies statistically significant patterns and generates trading signals. Triggers: "季节性", "日历效应", "月份效应", "周一效应", "年初效应", "节假日效应", "财报季效应", "时间模式", "季節性", "日曆效應", "月份效應", "周一效應", "年初效應", "節假日效應", "財報季效應", "seasonality", "calendar effect", "January effect", "day of week effect", "holiday effect", "earnings season effect", "seasonal pattern", "time series anomaly", "月度效应", "月度效應", "monthly seasonality".

longbridge-finance-query

Cross-market financial metrics batch query — revenue, net profit, ROE, debt ratio, free cash flow, gross margin for one or more symbols across HK / US / A-share / SG markets. Supports multi-symbol horizontal comparison, similar to natural-language financial screening. Triggers: "财务数据查询", "财务指标", "营收查询", "净利润查询", "ROE查询", "负债率", "现金流查询", "毛利率查询", "财务数据批量", "財務數據查詢", "財務指標", "營收查詢", "淨利潤查詢", "ROE查詢", "負債率", "現金流查詢", "毛利率查詢", "financial data query", "revenue query", "net profit query", "ROE query", "debt ratio", "free cash flow", "gross margin query", "financial metrics", "financial comparison", "batch financials".

longbridge-ownership

Share structure and shareholder query — total shares, circulating shares, restricted shares, top-10 shareholders (circulating and total), major institutional holders, controlling shareholder / beneficial owner relationships. Shareholder count and pledge data are not available via Longbridge; check filings for those. Triggers: "股本结构", "前十大股东", "流通股东", "实控人", "股权结构", "主要持股人", "大股东", "股东查询", "股本結構", "前十大股東", "流通股東", "實控人", "股權結構", "主要持股人", "大股東", "shareholder structure", "major shareholders", "top 10 shareholders", "controlling shareholder", "share structure", "institutional holders", "beneficial owner", "ownership structure", "free float".

project-setup

One-time project setup for creative writing. Interviews you about your project, collects writing samples, proposes kb structure, and creates AGENTS.md with project conventions.

pinme-email

Use this skill when a PinMe project (Worker TypeScript) needs to integrate email sending (send_email). Guides AI to generate correct Worker TS code.

Condition-Based Waiting

Replace arbitrary timeouts with condition polling for reliable async tests

docs

Create or refresh project documentation in `docs/`. Use when a project needs docs initialized or existing docs synced to the current codebase.

deep-analysis

Performs focused, depth-first investigation of specific reverse engineering questions through iterative analysis and database improvement. Answers questions like "What does this function do?", "Does this use crypto?", "What's the C2 address?", "Fix types in this function". Makes incremental improvements (renaming, retyping, commenting) to aid understanding. Returns evidence-based answers with new investigation threads. Use after binary-triage for investigating specific suspicious areas or when user asks focused questions about binary behavior.

luxury

High-end dark aesthetic with bold headings, monochromatic palette, and premium feel for luxury brand experiences.

skill-creator

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Codex's capabilities with specialized knowledge, workflows, or tool integrations.

longbridge-event-strategy

Event-driven investment strategy — identify and analyse corporate events (M&A, spinoffs, buybacks, index rebalancing, lockup expiry) that create pricing dislocations. Framework: event identification → sentiment scoring → historical price reaction → position sizing. Uses Longbridge news / filings / calendar data as signal inputs. Triggers: "事件驱动", "并购套利", "指数调整", "解禁套利", "事件策略", "公司事件策略", "事件投资", "套利机会", "事件驅動", "並購套利", "指數調整", "解禁套利", "事件策略", "公司事件策略", "event-driven", "event strategy", "merger arbitrage", "index rebalancing", "lockup expiry", "event investing", "corporate event trading", "special situation", "spinoff", "buyback catalyst".

longbridge-operating

Quarterly and annual operating data for listed companies via Longbridge Securities — revenue, net income, EPS, ROE, gross margin, and other KPIs broken down by reporting period. Complements longbridge-fundamental with period-by-period trend detail. Note: currently returns data for HK-listed stocks only; US and A-share symbols return empty results. Triggers: "经营数据", "运营指标", "分季度财务", "经营情况", "季度营收", "季度利润", "财务趋势", "经營數據", "運營指標", "分季度財務", "季度營收", "季度利潤", "operating data", "quarterly financials", "operating indicators", "quarterly revenue", "quarterly profit", "financial trend by period", "700.HK operating", "按期次财务", "按期次財務".

longbridge-ml-strategy

Machine-learning prediction strategy framework via Longbridge Securities — walk-forward rolling training with feature engineering (MACD, RSI, Bollinger Band width, volume change rate) and a scikit-learn classifier (Random Forest / Gradient Boosting); retrains every 60 days, predicts 5-day direction probability; evaluates win rate, profit factor, and Sharpe ratio. For reference only — not investment advice. Triggers: "机器学习", "ML策略", "预测模型", "随机森林", "梯度提升", "深度学习", "AI选股", "walk-forward", "機器學習", "ML策略", "預測模型", "隨機森林", "梯度提升", "machine learning", "ML strategy", "predictive model", "random forest", "gradient boosting", "AI stock selection", "walk-forward", "rolling training", "feature engineering", "scikit-learn", "XGBoost".

longbridge-valuation-rank

Industry valuation rank time series for a single stock via Longbridge — tracks how a stock's PE / PB / PS / dividend-yield rank within its sector has changed over time (rank N of total M). Answers "is my stock becoming relatively cheaper or more expensive vs peers?" Complements longbridge-valuation (single-stock percentile history) and longbridge-industry-valuation (current sector snapshot). Triggers: "行业排名变化", "估值排名", "PE排名历史", "行业估值位置", "排名走势", "估值相对同业", "行業排名變化", "估值排名", "PE排名歷史", "行業估值位置", "排名走勢", "valuation rank", "industry rank history", "PE rank trend", "relative valuation rank", "sector ranking over time", "how does AAPL rank in industry PE".

archive

Use when a Beat change is complete (implemented, or distilled and verified) and ready to archive — not for verifying implementation

plan

Use when breaking down a Beat change spec into tasks — not for spec creation or implementation

agent-observability-experiment-analyzer

Analyze LLM experiment results. Handles single or comparative experiments, exploratory or Q&A modes. Use when user says "analyze experiment", "compare experiments", "analyze against baseline", or provides one or two experiment IDs for analysis.

lifestyle-ad

Design lifestyle ads where product capability supports a desirable human state. Produce or review shot scripts with grounded product behavior and subtle proof.

claude-md-improver

Audit and improve CLAUDE.md files in repositories. Use when user asks to check, audit, update, improve, or fix CLAUDE.md files. Scans for all CLAUDE.md files, evaluates quality against templates, outputs quality report, then makes targeted updates. Also use when the user mentions "CLAUDE.md maintenance" or "project memory optimization".

uniswap-monitor-teneo

AI agent for Uniswap Monitor via Teneo Protocol. Use this skill when the user needs Uniswap Monitor via the bundled Teneo CLI and you need the live commands, arguments, or pricing before execution.

using-git-worktrees

Git worktrees create isolated workspaces sharing the same repository, allowing work on multiple branches simultaneously without switching.

draw-ui

设计 App、后台、游戏界面和完整网站落地页，生成 UI 设计稿；按需把设计稿或截图还原为 HTML/CSS、现有前端项目或微信小程序。用户要求页面设计、完整落地页、截图还原或小程序高保真时使用。不用于普通插画、海报、故事板、纯图片编辑或无设计任务的业务代码修改。

stakeholder-communication

Manage stakeholder expectations and engagement through targeted communication, regular updates, and relationship building. Tailor messaging for different stakeholder groups and priorities.

backend-testing

Turn backend test ambiguity into one practical backend test packet. Use when the user needs API/service/repository/auth-flow coverage design, fixture or seed/reset strategy, container-vs-mock dependency choices, contract/API compatibility checks, or flaky backend-suite stabilization across local and CI.

hf-cloud-sagemaker-iam-preflight

Ensure a usable SageMaker execution role exists before deploying or training. Use this skill whenever about to create a SageMaker endpoint, model, training job, or any resource that requires an execution role. Use it especially when the user has not provided a role ARN explicitly, when scripts are about to call `iam:CreateRole`, or when an AccessDenied error mentions an IAM action. Never blindly call `iam:CreateRole` — always check for existing roles first. This skill prevents the most common SageMaker deployment failure: trying to create IAM resources from an SSO principal that has no IAM write permissions.

mobile-ux-optimizer

Mobile-first UX optimization for touch interfaces, responsive layouts, and performance. Use for viewport handling, touch targets, gestures, mobile navigation. Activate on mobile, touch, responsive, dvh, viewport, safe area, hamburger menu. NOT for native app development (use React Native skills), desktop-only features, or general CSS (use Tailwind docs).

agent-observability-session-classify

Classify whether user intent was satisfied in a Datadog Agent Observability trace or session. Three modes: (1) session_id — classify a single CMD+I assistant session with RUM; (2) trace_id — classify a single Agent Observability trace without RUM; (3) ml_app — sample and classify multiple sessions or traces from a given LLM app. Output is compact by default (verdict + one-sentence reason). Use when evaluating satisfaction, classifying sessions/traces, labeling data, or generating signal for agent-observability-eval-pipeline or agent-observability-trace-rca.

wan-3-0-prime-reference-to-video

Build video clips from reference images, reference videos, and reference audio with Wan-AI Wan 3.0 Prime Reference to Video on RunComfy. Up to 10 reference images, 5 reference videos and 5 reference audio clips are bound to a prompt that names them as "Image 1", "Video 1", "Audio 1", giving character, product and scene consistency across a 2-30 second shot at 480p, 720p or 1080p with a synchronized audio track. Documents the full input schema, the counted-second pricing model (reference videos are billed as duration, images and audio are not), and when to route to Wan 3.0 Prime text-to-video / image-to-video, Wan 2.7 or Seedance 2.0 Pro instead. Calls `runcomfy run wan-ai/wan-3.0-prime/reference-to-video` through the local RunComfy CLI. Triggers on "wan 3 prime reference to video", "wan 3.0 prime", "wan3 prime", "reference to video", "ref2v", "keep the same character across shots", "video from reference images", or any explicit ask to generate video from references with this model.

seedance-2-5-reference-to-video

Generate reference-guided 1080p video with ByteDance Seedance 2.5 Reference to Video on RunComfy via the `runcomfy` CLI. Feed up to 9 reference images, 1-3 reference video clips, and 3 reference audio files into one call and get a 4-30 second 1080p clip with native synchronized audio, identity and style locked to your references. Documents the full input schema (images / videos / audios / aspect_ratio / duration / generate_audio), the counted-seconds billing model ($0.53 per second of reference video duration plus output duration), the 480p draft-then-deliver workflow, and when to route to Seedance 2.5 text-to-video, image-to-video, or Seedance 2.0 Pro instead. Calls `runcomfy run bytedance/seedance-2.5/reference-to-video/1080p`. Triggers on "seedance 2.5", "seedance 2.5 reference to video", "reference to video", "reference-to-video", "seedance 1080p", "ByteDance Seedance 2.5", "consistent character video", "style-locked video", or any explicit ask to generate video from reference images and clips.

readme-i18n

Use when the user wants to translate a repository README, make a repo multilingual, localize docs, add a language switcher, internationalize the README, or update localized README variants in a GitHub-style repository.

secure-linux-web-hosting

Use when setting up, hardening, or reviewing a cloud server for self-hosting, including DNS, SSH, firewalls, Nginx, static-site hosting, reverse-proxying an app, HTTPS with Let's Encrypt or ACME clients, safe HTTP-to-HTTPS redirects, or optional post-launch network tuning such as BBR.

opensource-guide-coach

Use when a user wants guidance on starting, contributing to, growing, governing, funding, securing, or sustaining an open source project, or asks about contributor onboarding, community health, maintainer burnout, code of conduct, metrics, legal basics, or open source project adoption.

anti-ui-slop

Stop coding agents from shipping generic UI. Use UIZZE's 800,000+ real web and iOS screens to build product-specific interfaces, define a design contract, cover required states, and run a hard finish gate. Use when designing, implementing, redesigning, critiquing, or pre-ship reviewing a web or iOS interface in Codex, Claude Code, Cursor, Copilot, or another coding agent. Trigger with "anti-ui-slop", "stop UI slop", "ground this UI in real screens", or "run the UI finish gate".

openclaw-secure-linux-cloud

Use when self-hosting OpenClaw on a cloud server, hardening a remote OpenClaw gateway, choosing between SSH tunneling, Tailscale, or reverse-proxy exposure, or reviewing Podman, pairing, sandboxing, token auth, and tool-permission defaults for a secure personal deployment.

use-my-browser

Use when work depends on the user's live browser session or visible rendered state rather than static fetches, especially for browser debugging contexts or DevTools-selected elements or requests, logged-in dashboards or CMS flows, localhost apps, forms, uploads, downloads, media inspection, DOM or iframe inspection, Shadow DOM, or browser failures that look like soft 404s, auth walls, anti-bot checks, or rate limits.

running-claude-code-via-litellm-copilot

Use when routing Claude Code through a local LiteLLM proxy to GitHub Copilot, reducing direct Anthropic spend, configuring ANTHROPIC_BASE_URL or ANTHROPIC_MODEL overrides, or troubleshooting Copilot proxy setup failures such as model-not-found, no localhost traffic, or GitHub 401/403 auth errors.

develop-userscripts

Use when building, debugging, packaging, or publishing browser userscripts for Tampermonkey or ScriptCat, including GM APIs, metadata blocks, permission issues, @match/@grant/@connect setup, ScriptCat background or scheduled scripts, UserConfig blocks, or subscription workflows.