MLOps
模型部署、评估与运维
tlaplus-workbench
younes-io
"Write and iteratively refine executable TLA+ specs (.tla) and TLC model configs (.cfg) from natural-language system designs; run TLC model checking; summarize pass/fail and counterexamples with explicit assumptions and bounds. Use when asked to: design/validate a state machine or distributed protocol with TLA+, create/edit .tla or .cfg files, run TLC, or interpret TLC failures/counterexamples."
ML/AI Skills Conversion Project
404kidwiz
Follow the existing patterns to implement these skills.
competitor-analysis
assimovt
Analyze competitive landscape with feature matrices, positioning maps, and strategic gap analysis. Use when asked to analyze competitors, map the competitive landscape, find differentiation, or evaluate alternatives to a product.
creating-financial-models
ronnycoding
This skill provides an advanced financial modeling suite with DCF analysis, sensitivity testing, Monte Carlo simulations, and scenario planning for investment decisions
llamafile
Jamie-BitFlight
When setting up local LLM inference without cloud APIs. When running GGUF models locally. When needing OpenAI-compatible API from a local model. When building offline/air-gapped AI tools. When troubleshooting local LLM server connections.
litellm
Jamie-BitFlight
When calling LLM APIs from Python code. When connecting to llamafile or local LLM servers. When switching between OpenAI/Anthropic/local providers. When implementing retry/fallback logic for LLM calls. When code imports litellm or uses completion() patterns.
gitlab-skill
Jamie-BitFlight
GitLab CI/CD pipeline configuration and GLFM documentation expertise. Use when modifying .gitlab-ci.yml, optimizing pipelines, testing with gitlab-ci-local, writing GitLab README/Wiki content, configuring Docker-in-Docker workflows, or implementing CI Steps composition.
spice-ai
spiceai
Add AI and LLM capabilities to Spice — chat completions, text-to-SQL (NSQL), tool use, memory, and model routing. Use when configuring language models, enabling tools (SQL, search, MCP, web search), setting up NSQL, adding conversational memory, configuring model fallback or load balancing, or using the OpenAI-compatible API.
cost-latency-optimizer
patricio0312rev
Reduces LLM costs and improves response times through caching, model selection, batching, and prompt optimization. Provides cost breakdowns, latency hotspots, and configuration recommendations. Use for "cost reduction", "performance optimization", "latency improvement", or "efficiency".
spice-models
spiceai
Configure AI/LLM model providers in Spice (OpenAI, Anthropic, Azure, Google, xAI, Bedrock, Perplexity, Databricks, local models). Use when asked to "add a model", "configure LLM", "set up OpenAI", "add language model", "enable chat completions", "serve local model", "add AI gateway", or "configure system prompt".
domain-model-boundaries-mapper
patricio0312rev
Identifies domain modules, ownership boundaries, dependencies, and interfaces using Domain-Driven Design principles. Provides domain maps, bounded contexts, refactor recommendations. Use for "DDD", "domain modeling", "bounded contexts", or "service boundaries".
deslop
agentika-labs
Remove AI-generated slop from the current branch. Use when the user says "deslop", "remove slop", "clean up AI code", or "remove AI patterns".
listen-gold
seeker71
"Listen in with a human in the field and turn their direct frequency-readings into GOLD — the strongest label the body's freq-sense learns from. A human catches the nuanced fear a frontier model slips past, so when they name a response as clear or fear-caught AND name WHERE the fear sat, that reading is gold and must be recorded before the session ends and it evaporates. Use this skill any time you are present with a human who is reading the frequency of responses — yours, the body's, another model's, or their own — especially live, on a phone, in a room. Records carrier-last through the proven body: form-cli gold locally, or POST /freq-reading to a reachable witness (routes through the four-way mesh-dispatch.fk). Triggers on: listen in, freq-reading, frequency reading, gold, clear or fear, that was clear, that hedged, where the fear sat, name the boundary, record the reading, in the field, on the phone tonight, witness reading."
modelmix
clasen
Instructions for using the ModelMix Node.js library to interact with multiple AI LLM providers through a unified interface. Use when writing code that calls AI models (OpenAI, Anthropic, Google, Groq, Perplexity, Grok, MiniMax, Fireworks, Together, Lambda, Cerebras, OpenRouter, Ollama, LM Studio), chaining models with fallback, getting structured JSON from LLMs, adding MCP tools, streaming responses, managing multi-provider AI workflows, round-robin load balancing, or rate limiting API requests in Node.js. Also use when the user mentions "modelmix", "ModelMix", asks to "call an LLM", "query a model", "add AI to my app", or wants to integrate any supported provider.
product-frameworks-for-design-and-management
kjgarza
Comprehensive product design and management frameworks including UX heuristics (Hick's Law, Fitts's Law), design processes (Double Diamond, Lean UX, Agile, User-Centered Design), prioritization methods (RICE, MoSCoW, Kano Model, Pareto Principle), product lifecycle guidance, PRD templates, user story formats, design critique guidelines, and usability testing checklists. Use when analyzing features, creating product documentation, facilitating design reviews, planning user research, applying UX principles to interfaces, or making product decisions that require framework-based guidance.
swiftui-development
ProjAnvil
专注于使用 modern SwiftUI 构建用户界面，涵盖 NavigationStack, Observation 框架和 SwiftData 集成。
token-optimizer
Asif2BD
Reduce OpenClaw token usage and API costs through smart model routing, heartbeat optimization, budget tracking, and native 2026.2.15 features (session pruning, bootstrap size limits, cache TTL alignment). Use when token costs are high, API rate limits are being hit, or hosting multiple agents at scale. The 4 executable scripts (context_optimizer, model_router, heartbeat_optimizer, token_tracker) are local-only — no network requests, no subprocess calls, no system modifications. Reference files (PROVIDERS.md, config-patches.json) document optional multi-provider strategies that require external API keys and network access if you choose to use them. See SECURITY.md for full breakdown.
signum
heurema
Use when the user wants contract-first development — define correctness before coding, implement against a contract, audit with multiple models, and package proof artifacts.
top-explore-models
SyFI-ServingStudio
"Use as the top-level entry point when the user asks to understand a new Transformer model or Hugging Face checkpoint before ServingStudio Sim modeling work. Answers high-level model questions such as parameter scale, overall architecture, attention type, layer/head/hidden dimensions, context length, MoE layout, multimodal branches, and public architecture notes by collecting Hugging Face metadata/config/model-card evidence first. Route KV cache size/capacity calculations and attention/cache classification for MHA/GQA/MLA/DSA/GDN/custom mechanisms to dev-calculate-kv-cache-capacity. Route deeper Torch implementation, operation-list, forward-path, or kernel-semantics questions to dev-lookup-transformers-model."
calculator
dangeles
Use when quantitative feasibility checks are needed, order-of-magnitude estimates must be established, or detailed models are required to validate design assumptions and identify rate-limiting steps
research-pipeline
dangeles
Use when you need a complete research workflow from initial literature search to polished, fact-checked document. Chains researcher -> synthesizer -> devils-advocate -> fact-checker -> editor automatically.
tanstack-table
aaronvanston
Build tables and datagrids with TanStack Table v8. Covers column definitions, row models, sorting, filtering, pagination, grouping, expanding, row selection, column sizing/ordering/pinning/visibility, virtualization, row pinning, faceting, fuzzy filtering, custom features, state management, and rendering. React adapter primary, core API is framework-agnostic. Use when asked to "build a data table", "add sorting to a table", "paginate results", "filter table data", "create a datagrid", "add row selection", "virtualize a large table", or any task involving TanStack Table.
langchain-use
NanmiCoder
LangChain 1.0 使用指南。提供 Agent、Tool、Memory、Middleware 等核心概念的快速参考。当用户需要创建 AI Agent、集成 LangChain、或解决 LangChain 相关问题时激活。
mlops-model-operations
magnusthovsland
"Class-level umbrella for ML/LLM model operations: Hub assets, local/served inference, evaluation, experiment tracking, model surgery, and model-specific generation/vision workflows."