MLOps
模型部署、评估与运维
fable-mode
mrtooher
Enforces staged execution discipline on large tasks: a written stage plan, delegation to named fable agents where the runtime supports it, a failable verification check at each stage, and a skeptical self-review before delivery. Trigger when the user explicitly asks ("do this thoroughly", "be systematic", "deep work mode") OR when the task objectively spans multiple files, multiple sources, or multiple sessions. Do NOT trigger on ordinary multi-step requests that a direct attempt handles fine. For a run pinned to a specific model, use fable-opus, fable-sonnet, or fable-haiku instead. Always-on guardrails (verify-before-flag, warning batching, sed safety) live in the companion execution-guardrails skill.
openrouter
rawwerks
Use this skill when the user wants to call different LLM models through OpenRouter's unified API, compare model responses, track costs and response times, or find the best model for a task. Triggers include requests to test models, benchmark performance, use specific providers (OpenAI, Anthropic, Google, etc.), or optimize for speed/cost.
firebase-ai-logic-basics
anoopithunt
Official skill for integrating Firebase AI Logic (Gemini API) into web applications. Covers setup, multimodal inference, structured output, and security.
agent-cost-optimization
arielperez82
Framework for analyzing agent telemetry data to optimize cost, token usage, and model tier selection. Use when tuning agent configurations, reviewing agent costs, selecting model tiers, or interpreting optimization_insights/cost_by_model/agent_usage_summary pipe outputs.
loom-nodejs
link-loom
Develop BACKEND Node.js services, APIs, routes, and models for Link Loom. Handles database interactions, business logic, and server-side operations using loom-sdk.
hqq-quantization
majiayu000
Half-Quadratic Quantization for LLMs without calibration data. Use when quantizing models to 4/3/2-bit precision without needing calibration datasets, for fast quantization workflows, or when deploying with vLLM or HuggingFace Transformers.
naval-almanack
kangarooking
当用户询问《纳瓦尔宝典》(The Almanack of Naval Ravikant) 相关内容，或在以下领域寻求建议时调用: 财富构建(怎么赚钱/股权期权/杠杆/副业与职业定位/时薪与外包决策)、判断与决策(重大选择纠结/心智模型/打工vs创业)、 幸福与心智(欲望与焦虑/接受与内耗/冥想/刷手机上瘾/愤怒与期望/攀比妒忌)、人生哲学(生命意义/该不该信玄学/诚实沟通/选朋友圈子/怎么读书学习)。 不适用于: 具体投资标的与选股、临床心理治疗、急性危机(先求助专业人士)。 Triggers: 纳瓦尔/Naval/纳瓦尔宝典/致富/财富/幸福/判断力/冥想/读书/人生意义
super-transcribe
ThePlasmak
"Unified speech-to-text skill. Use when the user asks to transcribe audio or video, generate subtitles, identify speakers, translate speech, search transcripts, diarize meetings, or perform any speech-to-text task. Also use when a voice message or audio file appears in chat and the user's intent to transcribe it is very clear."
senior-prompt-engineer
nimeshgurung
World-class prompt engineering skill for LLM optimization, prompt patterns, structured outputs, and AI product development. Expertise in Claude, GPT-4, prompt design patterns, few-shot learning, chain-of-thought, and AI evaluation. Includes RAG optimization, agent design, and LLM system architecture. Use when building AI products, optimizing LLM performance, designing agentic systems, or implementing advanced prompting techniques.
windows
hk-vk
Note: Qwen 3 coder is a 30B parameter model requiring at least 24GB of VRAM to run smoothly. More is required for longer context lengths.
Data Pipeline & Loaders
sovr610
This skill should be used when the user asks to "load a dataset", "create data loaders", "add data augmentation", "preprocess input data", "set up training data", "create validation splits", "register a dataset", "configure data pipeline", "handle multi-modal data loading", "load MNIST/CIFAR/ImageNet", "load event camera data", "load sequence data", "set up few-shot episodes", "create a dataset registry", or needs guidance on data loading, preprocessing, augmentation, or dataset management for the brain_ai 7-phase training pipeline.
ddd
CacaoRick
DDD 領域驅動設計 - 戰略設計。專注於產出 Domain Model 文件，包括識別核心域/支撐域/通用域、拆分 Bounded Context、定義 Ubiquitous Language 與 Context Map。Use when (1) the user asks how to split Bounded Contexts, (2) the user asks to define Ubiquitous Language, (3) the user wants to create a Domain Model document, (4) the project uses DDD architecture and requires strategic design.
ark-embedding
jneless
skill of vole engine ark model platform embedding api
two-sigma-ml-at-scale
copyleftdev
Build trading systems in the style of Two Sigma, the systematic investment manager pioneering machine learning at scale. Emphasizes alternative data, distributed computing, feature engineering, and rigorous ML infrastructure. Use when building ML pipelines for alpha research, feature stores, or large-scale backtesting systems.
ai-dev-workflow
CacaoRick
由文件串起完整的 AI 開發流程。涵蓋領域建模與程式碼組織（DDD）以及 AI 開發規範設定（Agent 規範）。Use when (1) the user asks to organize code by business features or define naming conventions, (2) creating or updating AGENTS.md / project rule files, (3) starting a new project from scratch, or (4) the agent needs the full development lifecycle.
cohere-typescript-sdk
RSHVR
Cohere TypeScript/JavaScript SDK reference for chat, streaming, embeddings, reranking, and tool use. Use when building Node.js or browser applications with Cohere APIs.
obsidian-knowledge-factory
vviniumeginn-del
Use when users need to clean, atomize, deduplicate, and archive mixed text files into an Obsidian vault using a local Python pipeline with an OpenAI-compatible API, without n8n.
mobile-ci-cd-evolution
Envy-7z
CI/CD optimization workflow for existing mobile pipelines (Android and Flutter) across Bitbucket, GitHub Actions, GitLab, Jenkins, and Fastlane. Use when auditing current CI/CD setup, reducing build time/flakiness, improving release safety, or planning incremental pipeline upgrades without destabilizing delivery.
Evaluation & Benchmarks
sovr610
This skill should be used when the user asks to "evaluate the model", "run benchmarks", "compute metrics", "measure accuracy", "test on MNIST", "compute F1 score", "generate confusion matrix", "evaluate few-shot", "measure anomaly detection", "run cognitive benchmarks", "compare model variants", "create evaluation report", "set up eval harness", or needs guidance on evaluation protocols, metrics computation, benchmark harnesses, or performance reporting for the brain_ai system.
compress-session
pedrohcgs
Distill the current conversation into a structured note (decisions made, open questions, file pointers with line numbers, next 1–3 actions) and save to quality_reports/session_logs/ before auto-compression. Differs from /checkpoint (explicit stop-point snapshot) and from auto-compaction (which truncates rather than distills). Use when context is approaching auto-compact threshold, when a long pipeline has accumulated many decisions, or when the user says "compress", "distil this session", "before we hit auto-compact", "structured handoff before context resets".
Odoo 18 Development Skills
hellbellies
```
analyze-function
carsten-j
Analyze the specified function from the given file with detailed line-by-line breakdown. Use when you need to understand what a function does, how it works, and why it was implemented that way.
status
pmco23
Use at any time to check pipeline state and get next-step guidance. When no pipeline is active, shows available workflow options and paths. No gate — always available.
debate
aroc
Orchestrate a debate between Claude and Codex to reach consensus. Use when the user asks for a "second opinion", "debate", "consensus", "crosscheck", or wants another model to review a proposal or diff.