majiayu000
- 75k skills
- 0 followers
- 735 repo stars
- 16 hours ago last updated
- ▌ Skill Checkup · majiayu000Check the agent skills installed on this machine across Claude Code and Codex for problems - broken links or files, same-name copies that drifted apart, descriptions that compete or miss requests their skill already handles, and skills loaded by both agents but used by only one. Use when the user asks to check, audit, clean up or tidy their skills, compare Claude Code and Codex skills, or asks why a skill did not trigger. For unused Claude Code skills and their token cost alone, /skill-doctor is enough.
- ▌ Code Film · majiayu000 bundle用纯代码(three.js 离线渲染 + 合成音频)做有故事、有电影质感的动画短片并导出 MP4。当用户想做短片、动画片、宣传片、角色动画、循环视频、片头、产品故事片,或者说“用代码做个视频”“做一个有趣的短片”“帮我设计个小动画”“make a short film / animated video / motion piece with code”时使用,即使没提 three.js。也用于继续修改这类片子。写实真人、实拍感、AI 视频生成、普通视频剪辑不适用。
- ▌ Repo Film · majiayu000 bundle把一个 git 仓库在一段时间里的真实提交做成 15–25 秒的炫酷短片:主线和各条分支按真实分叉、合并同时生长成流程图,越长越快再拉远看全景;仓库有 you-are-here 项目地图时,整张网会收拢成地图里的阻塞链,再沿地图讲卡点、当前位置和一句话总结,最后接可改的结尾卡片。一行命令出 MP4,基于 motion-stack 渲染器。当用户说「把这个仓库这周 / 昨晚做的事做成视频」「做个项目进展短片」「git 记录做成流程图动画」「给 you-are-here / 某项目做宣传视频」「像上次 Keepline 那个视频一样做一个」时使用。只在明确请求时触发,忽略顺带提及。
- ▌ Loop Close · majiayu000 bundle闭环推进一件事(功能、调研、文档、增长活动、开源项目):定目标与失败判定 → 走通任务 → 定规则 → 实现 → 验收 → 发布触达 → 看数据 → 复盘决策,结论回写。用户说「闭环」「功能卡」「复盘」「这件事怎么做完」时使用。一次性小改动、单纯写 PRD(prd-master)、只做指标设计(product-analytics)不触发。 Use when explicitly requested; ignore mentions/traces.
- ▌ Form Motion · majiayu000 bundleFORM UI 动效词典:10 个克制、跟手、可访问的界面微交互,每个都配原始提示词、可改项、验收标准和自动验收脚本——连续形变按钮、液体标签指示器、弹性滑块、柔性开关、原位展开卡片、层叠通知、命令面板、拖动排序、跟随提示、自绘折线图。共用一个从当前位置和速度重定向的阻尼弹簧。当用户说「按钮自己变成加载和完成」「选中条像软糖追过去」「拖到头还能拽一点」「开关圆点拉长」「卡片原位展开」「通知层叠」「Cmd+K 命令面板」「拖动排序」「提示跟随鼠标」「折线自己画出来」,或点名 FORM 动效库、要把某个样例用到自己的页面时使用。只在明确请求时触发,忽略顺带提及。
- ▌ Motion Stack · majiayu000 bundle用代码做动效短片的通用底座:按 brief 从 20 个开源库(Three.js、PixiJS、p5、Paper、Matter、Rapier、anime.js、tsParticles、OGL、GL Transitions、Tone.js、Chatterbox、faster-whisper、Remotion、HyperFrames、FFmpeg 等)里挑最小组合,写成「画面是时间的纯函数」的场景,用本地无头渲染器出 MP4(可带离线合成的音效),并自动写出可复现清单(库版本、字体、文件哈希、引擎版本、命令)。当用户说「做个动效 / motion 短片 / 物理动画 / 3D 动画视频 / 用 three.js 做个视频 / 用这套 motion stack」时使用。MV、知识视频、尺寸图动画走各自专用 skill。只在明确请求时触发,忽略顺带提及。
- ▌ Code Built 3d · majiayu000 bundle用代码从零件搭 3D 场景并能反过来拆开:每个物件是一个往 Assembly 里加零件的函数(盒子、圆柱、球、线),渲染成技术插画(白灰三档面、细黑描边、白雾),正交等距相机在镜头之间连续运镜,外壳可以降下去露出内部,选中的部分能爆炸成零件并列出零件清单,场景按块无限生成。零图片、零外部资源,three.js 已内置。演示是「无尽楼」:向上无尽的公寓楼,电梯一直往上走,点任意一层进去看、拆开。当用户说「用代码搭一个 3D 场景」「等距 3D 城市」「做成 Train of Thoughts 那种」「全程代码生成不要图片」「能剖开 / 去壳 / 爆炸的 3D」「无限生成的场景」「three.js 做构造」时使用。只在明确请求时触发,忽略顺带提及。
- ▌ Hud Interface · majiayu000 bundle游戏 HUD 式界面:黑白纸墨加一个薄荷强调色,斜切黑标签、空心巨型数字、粗黑汉字标题、环形转动文字、歪着摆的纸卡片、黑色底栏大斜体计数和条形表,背后是流动的液体大理石着色器;可选用 HTML-in-Canvas 把真实 DOM 面板画进 WebGL 再做玻璃折射和色散(Chrome 实验开关,不支持时自动退回普通 DOM)。带噪音档位 quiet / normal / loud,默认 quiet。适合拆解台、零件清单、机器人或任务控制台这类要有「仪表感」的页面。当用户说「游戏 HUD 风格」「做成仪表板那种酷炫界面」「HTML-in-Canvas」「把 DOM 画进 WebGL」「液体大理石背景」「斜切标签 / 巨型数字 / 环形文字」时使用。只在明确请求时触发,忽略顺带提及。
- ▌ Ascii Backdrop · majiayu000 bundleTurn any subject into an animated ASCII backdrop behind a UI, like a dark prompt page with jellyfish drawn in drifting characters, or a swarm of colour-tile bodies that veer away from the cursor. The subject is open-ended: creatures, objects, particles, smoke, a word or logo, an image or video, a per-pixel field, a shaded 3D solid, or your own canvas draw function. Six print modes can be mixed: brightness glyphs, edge-following glyphs, shape-matched glyphs, braille dots, ordered-dither dots and colour tiles; plus cursor glow, click ripples, a CRT terminal finish, and two page helpers (text decode and image-to-ASCII dissolve). Ships a dependency-free JS engine with HTML and React wiring. Use when the user asks for an ASCII background, ASCII art animation, glyph or braille backdrop, text scramble/decode, or an ASCII image reveal, or says 'ASCII 背景' / '字符画动效'.
- ▌ Avoidance Coach · majiayu000 bundleGuide a self-help conversation for someone stuck in an avoidance loop: low mood, emptiness, stress, or loneliness they escape from with games, scrolling, or other distractions instead of facing. Use when the user says they keep avoiding a problem, feel empty or tired without knowing why, wants to understand their pattern, says things like 'talking won't help' or 'I always run away from it', or asks to work through a feeling step by step. Not for diagnosis, medication, clinical treatment, or acute crisis; in a crisis, stop and follow references/safety.md.
- ▌ Agent QA Testing · majiayu000Use when a project uses agent-qa and the user asks to validate its YAML, run a selected regression test, or triage a failed/healed run with memidex evidence. For a general app feature inventory, use app-user-story-qa.
- ▌ Blender Editable 3d · majiayu000 bundle根据参考图、带尺寸的设计图或文字概念,在 Blender 中制作由独立命名零件组成的可编辑 3D 模型,再按用户要求修改尺寸、部件和配色,交付 .blend、GLB 和多角度渲染。用于图纸转三维、产品概念建模和风格化硬表面道具,以及“做一个同样的案例,换成坦克或其他物体”等请求;不用于人物角色、头发或脸部建模,这些请求使用 blender-reference-to-3d。以可继续修改的几何工程为核心;不默认制作拆解动画、录屏、绑定或制造模型。
- ▌ Illustrated Gallery · majiayu000 bundleArt-direct complete illustrated websites and write project-specific image-generation prompts. Use when the user requests Illustrated Gallery, a gallery-inspired visual style, or $illustrated-gallery. Covers visual direction, original artwork, and page-wide composition; not a fixed website clone, standalone image task, or generic dashboard skill.
- ▌ Blender Reference To 3d · majiayu000 bundle根据人物参考图在 Blender 制作、修整和配色可编辑角色,尤其适用于二次元头部与头发、静态多角度展示。覆盖输入拆解、轮廓与五官标注、九视图参考、模板选型、几何实现、材质配色、视觉检查和工程交付;当用户要求图转三维头像、按参考建模、修复相似度、完善配色或运行完整人物制作流程时使用。仅分析、写流程或封装技能时不修改模型。
- ▌ Model Inference Optimize · majiayu000 bundle优化实际模型推理链路,将正确性对齐、分段 profiling、显存与数据搬运、TensorRT/ONNX/PyTorch 后端、attention/kernel、FP8/compile、缓存与少步采样、质量回归、GPU 成本和服务验收串成同一实验闭环。当用户要求推理提速、降低显存或 GPU 成本、复现模型效果、定位 GPU 利用率低、优化图像/视频/扩散模型或自托管 LLM 时使用,提供 model inference performance optimization 的脱敏案例。纯 API 调用、提示词或 token 用量优化,以及电脑卡顿,不进入本技能的模型推理实验流程。
- ▌ Rate Limit Policy Design · majiayu000 bundleDesign rate limits using identity scope, burst behavior, distributed enforcement, and actionable retry responses.
- ▌ Domain Agentic Context · majiayu000 bundleSpecialized agentic and context engineering prompt methodology for Claude. Use when designing AI agents — system prompts, tool definitions, context window architectures, failure mode planning, multi-agent coordination, agent evaluation. Trigger on: "design an agent," "build a system prompt for an agent," "architect an agent's context," "design tool definitions," "agent system prompt," "context window architecture," "tool interface design," "evaluate agent behavior," "multi-agent coordination," "context engineering for agents," "ACI design." Provides 12 tested approaches for agent architecture (designing agents, not using them). Do NOT use for evaluating or scoring individual non-agent prompts (use prompt-validation if available). Do NOT use for general software architecture that hosts agents (use standard technical reasoning for that).
- ▌ AI Vendor Due Diligence · majiayu000 bundleUse this skill when the company is about to buy, renew or switch on an AI product or an AI feature inside an existing product, and asks "assess this AI vendor", "do they train on our data", "AI supplier questionnaire", "review the AI terms", "can we use this model provider with customer data", "who are the subprocessors behind this AI tool" or "AI vendor risk assessment". Sizes the review to the use, reads the terms, data processing agreement and documentation before asking questions, checks training use, retention, model changes, subprocessors, transfers, security and exit, and writes an assessment with a recommendation plus the follow-up questions for the vendor.
- ▌ AI Ml Engineering · majiayu000 bundleAI/机器学习工程全流程技能。覆盖GPU加速(RAPIDS/CuPy/Numba/CUDA kernel调优)、scikit-learn经典ML、SHAP模型解释、特征工程与数据泄漏防护、时序验证、PyTorch Lightning分布式训练、强化学习(Stable-Baselines3)、TimesFM零样本预测、EDA数据体检、pymoo多目标优化。触发词:机器学习、深度学习、GPU加速、CUDA、kernel调优、模型训练、特征工程、数据泄漏、模型解释、SHAP、sklearn、交叉验证、分布式训练、强化学习、时序预测、机器学习工程, machine learning, GPU acceleration, CUDA kernel, feature engineering, data leakage, SHAP, scikit-learn, cross validation, distributed training, PyTorch Lightning, reinforcement learning, time series forecasting, TimesFM, EDA, optimization。用于:加速数据管线、训练/调参/解释模型、检测特征泄漏、配置分布式训练、编写CUDA kernel等场景。
- ▌ Looot Google Maps Reviews · majiayu000 bundleRead Google Maps reviews. Use when the user says things like "get the reviews for this Google Maps place", "what do customers complain about at this restaurant" or "pull the latest reviews with owner replies". Runs through looot, billed per call.
- ▌ Service De Waba Integrate · majiayu000 bundleConfirm the customer's WhatsApp Business Account is shared with Salesforce on Meta's side. TRIGGER when the user asks to link, share, or confirm sharing of a WhatsApp Business Account with Salesforce, or to verify Salesforce appears in the WABA's Partners list. Pure instruction + one user-confirmation prompt — no API calls, no token handling. Sharing happens in Meta Business Suite (UI-only; Meta has no Graph endpoint for this). Renders step-by-step instructions with Salesforce's Business Manager ID substituted in, then asks the user to confirm whether "Salesforce" appears in the WABA's Partners list. DO NOT TRIGGER when the user needs full end-to-end WhatsApp setup (use `service-de-headless-channel-configure`) or to insert the channel record itself (use `service-de-channel-create`).
- ▌ Agent System Output Hygiene Ops · majiayu000 bundleAudit and remediate AgentSystem cron output hygiene by detecting raw tool logs, stack traces, and noisy summaries. Use when jobs leak 403/429/internal details, summaries are not user-safe, or post-change validation is required.
- ▌ Multiagent Memory Setup · majiayu000 bundleImport the user's own Claude Code and/or Codex session history (the prompts they typed and the agents' answers) into one cognee dataset, then answer questions over it, across every agent and per agent. Use when someone wants to load, import or backfill their previous coding-agent sessions into cognee, give several coding agents one shared memory of past work, or ask what their earlier sessions were about. Runs locally; reads ~/.claude and ~/.codex read-only, writes only to cognee's memory.
- ▌ Looot Company Email Format · majiayu000 bundleFind a company's email format. Use when the user says things like "what is the email format at example.com", "how are emails written at this company" or "guess the email pattern for example.com". Runs through looot, billed per call.
- ▌ Run Shopify App Research · majiayu000 bundleRun an Applora research workflow from a prepared brief or a direct Shopify App Store question. Use for selected templates, business context, output language, prior reports, saving findings or continuing research; specialized skills handle deeper domain analysis.
- ▌ Glycowork · majiayu000 bundleGlycan data science with the glycowork Python package. Use for any code that touches glycans, oligosaccharides, polysaccharides or glycoconjugates, including parsing and converting glycan notations (IUPAC-condensed/extended, WURCS, GlycoCT, GlyTouCan IDs, Oxford, LinearCode, GLYCAM, KCF, SMILES), compositions and masses, m/z to composition to structure, motif and epitope search (Lewis, sialyl, core fucose, LacNAc, blood groups), SNFG drawing, differential glycomics, glycoproteomics and lectin-array statistics, glycan databases with species/tissue/disease labels, protein-glycan binding data, biosynthetic networks, and glycan machine learning. Use it instead of writing custom glycan parsers, monosaccharide dictionaries, mass tables or string matching.
- ▌ Sovereign Intelligence Evolution Sie · majiayu000 bundleHabilidade especializada em converter dados operacionais (logs, alertas, feedbacks) em conhecimento estratégico por meio de sessões de 'Self-Study'.
- ▌ Looot Local Business Search · majiayu000 bundleFind local businesses. Use when the user says things like "find coffee shops in Austin", "list dentists near this address with their phone numbers" or "build a list of gyms in Lisbon". Runs through looot, billed per call.
- ▌ Reachai Capability Retrofit · majiayu000 bundleRetrofit an already-connected business system so ReachAI can operate more of it. Inventory business operations that fit the developer's goal, let the developer pick which ones to open, declare the picked ones with the ReachAI SDK (Java, C#, Python, Node.js, Go or PHP), trigger one audited SDK sync, and report. Use when asked to "let AI do X in our system", open more business methods to ReachAI, or run a ReachAI CAPABILITY_RETROFIT task.
- ▌ Huawei Cloud Cost Estimation · majiayu000 bundleEstimate Huawei Cloud pre-order costs, provision allowlisted resources through hcloud, and guide unsubscribe requests through the console. Use for Huawei Cloud pricing, budgets, resource creation, unsubscribe guidance, or hcloud readiness checks.
- ▌ Deslop 8 · majiayu000 bundleRemove AI-generated code slop that a branch added and match the code around it.
- ▌ Reproducible Ml Experiments · majiayu000 bundleImplement, run, and compare ML experiments with traceable configurations and artifacts. Use when training, fine-tuning, evaluating, or reproducing a model.
- ▌ Ak Vault · majiayu000 bundleStore, find, sync, rotate, and use secrets (API keys, tokens, .env values) through AgentKit's `ak secrets` vault without ever printing a value, while honoring a user's recorded choice of .env files or another secret manager. Use when a task touches a credential, a .env.* file, or a deploy or API call that needs a key.
- ▌ Check 5 · majiayu000 bundleRuns the ai-toolkit pre-PR checks that AGENTS.md requires for the files changed on this branch, and lists the hand-written per-agent files that may need mirroring. Run it from the ai-toolkit checkout before opening a PR. Never runs on its own.
- ▌ Dataset Integrity Assessment · majiayu000 bundleAudits dataset sources, labels, transformations, review, versioning, split integrity, retraining triggers, and rollback controls for poisoning and targeted degradation risk. Use when reviewing an ML training or evaluation data supply chain.
- ▌ Looot Find Professional Emails · majiayu000 bundleFind a professional email. Use when the user says things like "find the work email of Jane Doe at example.com", "get me a business email for this person" or "who can I email at example.com". Runs through looot, billed per call.
- ▌ Investigate And Fact Check · majiayu000 bundleRequires a compatible AI agent and web access for live checks; optional Python 3.10+ case tools. Trace claims and media to primary sources, test counterevidence, and preserve evidence with explicit uncertainty. Optional public-source adapters use crt.sh or GitHub.
- ▌ Mureo Meta Ads · majiayu000 bundleMeta Ads: Manage campaigns, ad sets, ads, insights, and audiences on Facebook/Instagram.
- ▌ Huawei Cloud Cost Estimation Cn · majiayu000 bundle估算华为云下单前费用,通过 hcloud 开通白名单内资源,并引导用户通过控制台处理退订。 适用于华为云询价、预算、资源创建、退订指引或 hcloud 可用性检查。
- ▌
- ▌ Nonprofit Fundraising Configure · majiayu000 bundleConfigure Nonprofit Cloud Fundraising on an org headlessly — assign permission sets/groups org-wide, enable Fundraising, create a default Gift Designation, configure gift/recurring-donation-failure settings and feature toggles, and offer optional enhanced features (Interest Tags, Interaction Summaries, Record Alerts, Life Events, Timeline). Use when the user asks to set up or enable Nonprofit Cloud Fundraising, configure gift/donation settings, provision an org for fundraising, assign Fundraising permission sets/groups, or create a default gift designation. Also: turn on Fundraising, set up recurring donations, prep an org for testing. Toggle writes use headless-360 operations, never the `IndustriesSettings` Metadata API; the Person Account profile-visibility fix is the one Profile metadata write. DO NOT TRIGGER when: enabling Person Accounts on its own (this skill's Layer 0 gate only), creating gift/donation records post-configuration, or configuring NPSP/Nonprofit Success Pack.
- ▌ Ml Decision Boundary Assessment · majiayu000 bundleEvaluates classifiers and scoring models used as security gates for adversarial robustness, preprocessing differences, thresholds, abstention, and fallback behavior. Use when an authorized assessment must measure whether bounded input changes alter a security decision.
- ▌ Competitive Scan · majiayu000 bundleScan competitor activity using auction insights and market signals. Use when the user asks about competitors, market dynamics, impression share changes, competitor moves, or competitive positioning. Also use when the user asks in Japanese (競合の動きを調べて / 競合分析 / インプレッションシェアの変化を確認 / 競合にシェアを取られていないか).
- ▌ Mlops 3 · majiayu000 bundleMLOps — Kubeflow 로 ML 파이프라인을 구성하고 KServe 로 서빙, MLOps 와 LLMOps 비교. Use when 학습부터 배포까지 ML 파이프라인을 설계할 때, 실험 추적은 mlops-tracking 을 쓴다.
- ▌
- ▌ History Search · majiayu000 bundleUse when you need to recall past work - before saying I don't know, what did we decide about X, has anyone seen this error, last time we, do you remember, why did we choose Z, how did we fix this before, what did agent X say or do about Y, search the history of one agent, a group of agents or every agent on the box. Dispatches the cheap history-search sub-agent over thrum transcript search and returns a short cited answer or NOT FOUND.
- ▌ Auto Improve Strict Shell Gate Install · majiayu000 bundleInstall the strict-shell-gate add-on of the agentic-auto-improve plugin in this project: a hook that reads each shell command before it runs and blocks a write into a source file that no approved contract covers. Says plainly what it gets wrong. Previews the files and the settings.json change, asks for one confirmation, then writes. The name carries the plugin name because you may have several plugins installed. Use when: user says /auto-improve-strict-shell-gate-install, 'install the strict shell gate', 'block shell writes before they happen', or 'turn on bash-gate'. Never overwrites a file that differs and never removes a row.
- ▌ Transcript Search 2 · majiayu000 bundleUse when you need to search past sessions, find what we decided about a topic, recall an earlier conversation, find where an agent discussed something, search agent history or transcripts, or work out which session mentioned a word, error or command. Covers thrum transcript search, show and reindex - count first, narrow, search with a small limit, read one window, delegate broad questions to a cheap sub-agent.
- ▌ Multicam Sync · majiayu000 bundleAI Multicam Editor synchronizes footage from a multi-camera / multi-recorder shoot — several cameras plus separate audio recorders covering one session, imported as loose clips — and makes a speaker-follow draft by default unless the user explicitly asks for sync only. Use when a user drops in multiple clips from the same recording and wants them aligned, asks for multicam / 多机位 / multi-angle sync, wants a "cut to whoever is talking" edit, or refers to camera A/B, angles, or separate lav or field recordings that need to line up with picture.
- ▌ Asset Import 2 · majiayu000 bundleImport local or downloaded media, or relink missing media to an existing ChatCut asset, through the hosted Claude plugin. Use the editor loopback bridge when the editor and files are on the same machine; use the upload helper for imports otherwise. Desktop MCP sessions use their own built-in import tools.
- ▌ Caption Burn · majiayu000 bundleBurned-in captions for a finished vertical video, three kinds. transcribe.py gets word timings from the video's own audio through the GooseWorks proxy (fal Whisper, charged to your GooseWorks credits, cents); captions.py burns one to three words at a time with Pillow + ffmpeg (no libass needed), either pinned to a split-screen seam (plate 25% above / 75% below) or at a fixed height, in a plate, outline or one-word serif style, with an optional red hook card; plates.py burns per-beat caption blocks (black, one union silhouette, placed in the emptiest band) for formats with no voice. The last caption (the CTA) holds to the final frame. Use as the last step of any video ad.
- ▌ Cco Review · majiayu000 bundleUse for /c-level-agents:cco-review, or to challenge a plan touching retention, segmentation or CS headcount.
- ▌
- ▌ Acceptance Design · majiayu000 bundleDefine executable acceptance for a bounded task before work or dispatch. Use when writing a worker brief, selecting success checks, or repairing an acceptance contract that cannot test the intended outcome.
- ▌ Auto Improve Strict Shell Gate Uninstall · majiayu000 bundleRemove the strict-shell-gate add-on of the agentic-auto-improve plugin from this project: the settings.json row of the bash-gate hook (PreToolUse on Bash), then the file .claude/hooks/bash-gate.py. Previews the row and the file, asks for one confirmation, then writes: the row first with a backup, then the file. Keeps the file when you edited it, and keeps the shared helpers. The add-on has no kill switch, so this skill is the way out. The name carries the plugin name because you may have several plugins installed. Use when: user says /auto-improve-strict-shell-gate-uninstall, 'remove the strict shell gate', 'turn off bash-gate', or 'uninstall the shell write gate'. Never deletes a helper.
- ▌ Spring AI 2 · majiayu000 bundleSpring AI 통합 — ChatClient, Advisors API, Structured Output, Function·Tool Calling, MCP, RAG, 비용 최적화. Use when Spring Boot 앱에 LLM 챗·RAG·MCP 기능을 붙이거나 Spring AI 구조를 설계할 때.
- ▌ Lego Show · majiayu000 bundleShows a model as a lego set: baseplate, bricks, joints, open studs, parts bin and suggested bricks, in text, an HTML picture, or a Mermaid diagram.
- ▌ Nowcast · majiayu000 bundleBuilds and runs a nowcasting system with release calendar, news decomposition and real-time evaluation.
- ▌ Adversarial Critique · majiayu000 bundleUse when an implementer is blocked mid-work by a design flaw, contradiction, or unforeseen fork with 2-3 ways to resolve it - produces a defended pick that unblocks the work and leaves an audit trail so the decision can be revisited if it later proves wrong
- ▌ AI Audit · majiayu000 bundleAudits AI/ML/LLM components: benchmarks, leakage, calibration, explainability, bias, red-team, confidentiality, monitoring.
- ▌ Aidlc Express · majiayu000 bundleRun the AI-DLC workflow with the express scope baked in — no scope detection. Lightest run: requirements to deploy, no design pass, no reviewers. Packaging over `/aidlc --scope express`, which works without this skill.
- ▌ Aidlc Feature · majiayu000 bundleRun the AI-DLC workflow with the feature scope baked in — no scope detection. Full lifecycle for new features, practical depth. Packaging over `/aidlc --scope feature`, which works without this skill.
- ▌ API Design Review · majiayu000 bundleDesign or review HTTP, RPC, or library APIs for stable contracts and predictable behavior. Use when adding endpoints or changing public interfaces.
- ▌ Lego Detach · majiayu000 bundleDetaches a brick from a model and keeps it in the parts bin. Inputs that depended on it switch to their declared stubs so the model keeps running.
- ▌ Model Spec · majiayu000 bundleWrites the model specification: equations, timing, closures, steady state, shocks, observables and known limitations.
- ▌ Call A Person · majiayu000 bundleRing a person's phone (the AI Maestro app) from an agent and see whether they answered. Use when a person must act now and a message would wait: a decision blocks you, something is breaking, or the operator asked to be called. Not for news, progress or anything a message can carry; and there is no audio yet, so a call says who is calling and the person answers or declines.
- ▌ Qwen Edit · majiayu000 bundleAI image editing prompting patterns for Qwen-Image (Qwen-Image-Edit-2511 by default, Qwen-Image-2.1 as a non-commercial Modal opt-in). Use when editing photos while preserving identity, reframing cropped images, changing clothing or accessories, adjusting poses, applying style transfers, or character transformations. Provides prompt patterns, parameter tuning, and examples.
- ▌ Audio Extract 3 · majiayu000 bundleExtract a 16kHz mono WAV audio track from any media file, probe metadata, and transcribe speech to text via the harness-provided LiteLLM sidecar (no internet required from inside the agent container).
- ▌ Blender 2 · majiayu000 bundleHeadless Blender (bpy) for 3D modelling and GLB export, with no native Blender install - build or edit a scene in Python, import OBJ/FBX/STL/PLY/glTF/.blend, export .glb and other formats, all through the blender-wasm MCP server and the pi-blender CLI. Use for "model a 3D object", "make a .glb of this", "convert this OBJ/FBX to glTF", "inspect this 3D model".
- ▌ Asc Workflow · majiayu000 bundleUse when defining, validating, running, resuming or auditing repo-local multi-step automations with asc workflow and .asc/workflow.json, including step outputs and safe release or TestFlight steps.
- ▌ Asc Aso Audit · majiayu000 bundleUse when running after pulling metadata with asc metadata pull. Run an offline ASO audit on canonical App Store metadata under ./metadata and surface keyword gaps using Astro MCP.
- ▌ Openapi Contract First · majiayu000 bundleContract-first OpenAPI workflow — author or update the spec, run the diff gate, regenerate types, write the controller. Use when adding or changing an HTTP endpoint.
- ▌ Testing API Contracts · majiayu000 bundleValidates API responses against OpenAPI or JSON Schema, detects breaking changes between spec versions, and builds consumer-driven contract checks. Use when an API has a published spec, when a backend change might break a client, when API tests assert only status codes, or when mocked fixtures need a guard against drifting from the real service.
- ▌ Ascend Cluster Fast Slow Rank Detector · majiayu000 bundle专门用于 Ascend 集群 Profiling 性能数据的“快慢卡”诊断专家技能。当用户提供【集群性能数据目录/路径】并要求分析【快慢卡】、【慢节点】、【负载不均衡】或【集群瓶颈】时,必须触发此技能。该技能会自动接收集群路径,调度相关工具输出快慢卡的宏观定性与微观根因(如 Host 下发瓶颈、算子计算劣化)。
- ▌ Spring Code Review Rubric · majiayu000 bundlePre-commit code review rubric for Spring Boot 4 changes. Used by `spring-code-reviewer` to produce `08-code-review.md` before any commit. Covers traceability, architecture, Spring idioms, error handling, data access, security, test quality, clarity, and migration.
- ▌ Testing LLM Features · majiayu000 bundleBuilds an offline eval suite for LLM-powered features: golden reference cases per prompt template, deterministic scoring where possible, a regression gate on prompt, model and retrieval changes, and CI wiring with promptfoo, Vitest or deepeval. Use when a product ships prompts, agents or RAG, when a model or prompt upgrade needs a regression check, when LLM output is currently verified by someone eyeballing it, or when asked how to test a feature whose output is non-deterministic.
- ▌ Consilium Principis · majiayu000 bundleA personal advisory board of AI characters modelled on real thinkers. Public-domain figures ship in the box (Sun Tzu, Marcus Aurelius, Epictetus, Machiavelli); the skill builds profiles and runs a session on the user's question, the advisers argue with each other and with the user, then synthesis and a next step. Quotes come only from verified sources.
- ▌ Cerebro Meet Analyst · majiayu000 bundleHabilidad para analizar transcripciones de reuniones del equipo del proyecto y extraerlas al OKF con estructura estandarizada. Incluye checklist obligatorio, Consulta Interactiva de Decisión ítem por ítem, detección de contradicciones con el cerebro digital y actualización automática de pendientes. Trigger: Cuando el usuario comparta un archivo de transcripción (.docx o .txt) de una reunión del equipo, o pida analizar/registrar una reunión.
- ▌ Store Analytics · majiayu000 bundleWooCommerce analytics and reporting. Track sales, analyze customer behavior, identify trends, and generate business insights.
- ▌ Chord · majiayu000 bundleRequires HTTPS access to api.getchord.xyz; optional Ask needs an account, API key and metered plan. Reads market data and builds unsigned DEX swaps; user-executed swaps incur network and route fees.
- ▌ Imla · majiayu000 bundleGlance sheet for a mock ML/AI system architecture interview at Google (staff level, 60 minutes), from the live transcript and screenshot on the connected cloud drive. Use when the user sends "imla".
- ▌ Imlc · majiayu000 bundleWhat to say and type in a mock ML/AI coding interview at Google (45 minutes, PyTorch or NumPy), from the live transcript and screenshot on the connected cloud drive. Use when the user sends "imlc".
- ▌ Bilan · majiayu000 bundleMeasure whether this session actually finished — the 0-10 completion number, computed from git, the carnet ledger, LIMITATION markers and CI instead of narrated. Use before reporting a completion number, when asked "where are we", at the end of a piece of work, and to find what a session that died left behind.
- ▌
- ▌ Imlas · majiayu000 bundleShort prose brief for a mock ML/AI system architecture interview at Google (staff level, 50 minutes), from the live transcript and screenshot on the connected cloud drive. Use when the user sends "imlas".
- ▌ 05 Foto · majiayu000 bundleÚsala cuando el estudiante pida la skill 05 o el requisito R7 (foto del medidor, permiso de cámara y conservación permanente de la imagen). Agrega la captura de foto al paso 3 del registro.
- ▌ 08 Envio · majiayu000 bundleÚsala cuando el estudiante pida la skill 08 o el requisito R9 (envío automático de lecturas pendientes al recuperar la conexión, sin duplicados, y banner de conexión y pendientes en todas las pantallas). También hace la verificación de lecturas registradas sin conexión (R4).
- ▌ Bud Dev · majiayu000 bundleHow the bud repo (a Neovim IDE driven by a small Go CLI) is built and extended. Use when working in a bud checkout, or when asked to add a language spec, a tool, a layout or a kit entry to bud.
- ▌ C2c Img · majiayu000 bundleGenerate or edit images with the user's local Codex (ChatGPT) image model, with no API key needed. Use it for ANY request to create a picture (photo, illustration, poster, banner, cover, wallpaper, avatar, icon, logo concept, sticker, mockup, meme) or to change an existing image or photo. That especially includes creative restyling that keeps the people and composition, such as turning a photo into cartoon, anime, pixel art, watercolor, sketch, oil painting, 3D clay or Ghibli style, as well as retouching, removing or replacing the background, and adding or removing objects. Prefer this skill over writing PIL, OpenCV or ImageMagick filter code for any creative or stylistic change. Use code only for exact mechanical edits (crop, resize, rotate, compress, convert format) or when the user explicitly asks for an algorithmic filter. Chinese triggers include 画/生成图片/生图/出图/海报/配图/头像/壁纸/改图/P图/抠图/换背景/卡通化/动漫化/漫画风/像素画/像素风/手绘风/水彩/素描/油画/吉卜力/Q版/风格化/再来一张. Don't use it for charts, diagrams or UI that should be built as code.
- ▌ Llmops · majiayu000 bundleLLMOps — RAG 운영, 프롬프트 버전 관리, Ragas 기반 LLM 평가, NeMo 가드레일. Use when LLM 앱을 운영 환경에서 품질 평가하고 프롬프트 변경을 관리하거나 가드레일을 붙일 때.
- ▌ Request · majiayu000 bundleFlag the open PR as ready for human review — request GitHub review from the two teammates who did not open it, then post a "ready for review" message to the het-core Slack channel with the PR title and link, tagging @channel. Use when the user wants to hand a finished PR off for review, or runs /request.
- ▌ Decisions · majiayu000 bundleZoom AI Services Decisions for typed predicates, constrained classification, ordered scoring, probability-aware routing, and inline image comparison. Use when building or debugging Decisions requests, production x-api-key authentication, standard/pro model selection, Boolean choices, or score-label normalization.
- ▌ Ad Agent · majiayu000 bundleDirect Creatify's Ad Agent MCP to make or edit a video ad end to end. It generates Boreal-H3 footage, keyframes, music and voice lines into a persistent project, you judge every clip from contact sheets, you assemble one index.html page, and it ships a gated MP4. Use it whenever the user wants to make a video ad with Creatify, compose a video from references, use the Ad Agent, resume a composer project, give notes on a take, regenerate shot N, edit this take, or change a cut, caption, CTA or music in a composed video. Requires the creatify-ad-agent MCP server (tools named composer_*); if those tools are missing, say the connector isn't connected instead of improvising.
- ▌ Castmate · majiayu000 bundleWhen you want to pull in the right AI model for the right job: hand a task to another agent CLI (Codex, Claude, Gemini, Grok, Amp, OpenCode, Cursor, Kiro, Antigravity), or send one prompt to several at once for independent opinions, then synthesize. Delegate mode gives one castmate the job, optionally with write access ('have codex write the migration'). Fan-out mode runs read-only reviews in parallel. Rounds mode lets castmates build on and challenge each other's findings. Calls the official CLIs you already have installed and logged in (Gemini CLI needs an API key on personal accounts); no MCP. Triggers on '/castmate,' 'ask codex,' 'have gemini do this,' 'get a second opinion,' 'what would other models say,' 'fan this out,' 'cross-check this with another model,' 'delegate this to codex.' Advisor personas for founder questions go to maker-council. Personal models, casting rules, and run outputs live in ~/.config/makerskills/castmate/.
- ▌ Chemdraw · majiayu000 bundleUse ChemDraw and cdxml-toolkit for editable chemical structures, reaction figures, ACS Document 1996 drawings for exams and papers, image reconstruction, scientific drawings, chemical analysis, and ChemDraw objects in Office; also install or diagnose their runtime.
- ▌ Aai Merkys · majiayu000 bundleUse aai-cli to ask Merkys questions and get grounded answers with cited sources, run deep research, and read its conversations, documents, data sources, and corpora.
- ▌ System One · majiayu000 bundleReference for System One decision models — Ollama's local POST /v1/systemone endpoint (nimble, clef, clef-flash) and TypeSafe's hosted Jev API (jev-latest / jev-1.13.0), plus the typesafe-sdk Python and @typesafe-ai/sdk JS clients. Covers the three question types (choice, score, noul), state shaping, structured instructions/criteria, confidence math, request limits and errors, the Ollama-vs-TypeSafe differences, Jev's known failure modes, and worked patterns from the official cookbooks (fan-out, confidence gating, composite scoring, intent routing, guardrails, RAG gating, extraction, re-ranking, hierarchical beam search). Use whenever the work mentions System One, systemone, Ollama decision models, nimble, clef, TypeSafe, Jev, noul, or classifying/scoring/yes-no judgments that code consumes instead of generated text.
- ▌ Agentcard · majiayu000 bundleLets an AI agent pay merchants securely with the user's own card — through Mastercard Agent Pay (the default for Mastercard holders, passkey-approved per purchase) or, for any other card, a one-time card number capped to the purchase — completes KYC verification when required, handles checkout 3DS challenges, manages the user's AgentCard agent email inbox, and uses the AgentCard USDC wallet for Base/x402 payments. Use when an agent needs to pay a merchant, complete required KYC, complete a checkout, fetch paid x402 content, send or read agent email, inspect spend limits, or troubleshoot AgentCard CLI payments.
- ▌ Msb Edit · majiayu000 bundleEditing an existing sandbox and the scriptable surface: msbctl edit's EDIT_SECTIONS each reusing the wizard's own pick_* function preselected from current state, set_toml_key() as the comment-preserving single-key editor (never rewrite a whole file), which settings are portable versus machine-local, the rebuild offer, _recover_broken() on an unparseable file, rename_sandbox() and move_project(), per-sandbox secrets written only through set_env_file_value()/try_set_env_file_value(), and the one-shot interfaces (ls/show --json, secret set, setup --claude-token). Use when changing msbctl edit, one of its sections, rename or move, secret storage, or a scripting interface.
- ▌ Finops AI · majiayu000 bundleAI 워크로드 비용 관리 — LLM 비용 구조, token economics·귀속, GPU 최적화, Shift-Left FinOps, model routing. Use when LLM API·GPU 비용이 늘어 토큰 단위로 추적하고 모델 라우팅으로 줄일 때.