WHEN: Working with mkosi configuration files, mkosi.conf, Include= directives. WHEN NOT: General Linux configuration, non-mkosi build systems.
پرامپت‌ساز تشریحی: پرامپت کامل با نقش، کانتکست، شرح کار، گاردریل و فرمت گزارش، برای وقتی محیط عجیب است یا قید سفت دارد (دسترسی، خط قرمز، لحن، فرمت خروجی دقیق).
پرامپت‌ساز هدف‌محور و سبک پیش‌فرض: به‌جای دیکته کردن مراحل، هدف را یک سطح بالاتر می‌گوید، به مدل راه تأیید کار خودش را می‌دهد و شرط توقف عینی می‌گذارد؛ برای کار چندسشنی خودِ…
Purge the context-mode knowledge base. Permanently deletes all indexed content and resets session stats. This is destructive and cannot be undone.
Digital marketing, paid traffic, growth, Meta Ads, Google Ads, funnels, copy, metrics, social media, email marketing, SEO
When the user wants to audit or optimize an App Store or Google Play listing. Also use when the user mentions 'ASO audit,' 'app store optimization,' 'optimize my app list — from…
When the user wants to reduce churn, build cancellation flows, set up save offers, recover failed payments, or implement retention strategies.
When the user wants help with pricing decisions, packaging, or monetization strategy. Also use when the user mentions 'pricing,' 'pricing tiers,' 'freemium,' 'free trial — from…
End-to-end methodology for AI agents and software engineers to add machine learning algorithms to existing non-ML codebases.
เริ่มโปรเจกต์ ML อย่างถูกวินัย — เลือก metric ให้ตรงราคาความผิดพลาดจริง, split ข้อมูลไม่ให้รั่ว (time-based / group-based), สร้าง baseline ที่โง่ที่สุดเป็นโค้ดจริงก่อน…
Establish heuristic and simple-model baselines that bound what complexity is worth. Use when starting any ML project or auditing whether a complex model earns its cost.
Evaluate ML data and models for leakage, splits, metrics, calibration, and errors.
Use when something is failing in ML/AI work — OOM, NaN, divergence, crashes, bad throughput, wrong outputs, dependency conflicts
Ml Dl Rl brain — domain knowledge compiled from raw sources into structured markdown pages with backlinks, tags, and source references.
Review model errors by hand, discover failing slices, and decide whether data or model fixes pay more.
Computer vision for buildings, image-to-floorplan, generative ML models, performance prediction, structural analysis ML, energy prediction, natural language to design, and point…
Master machine learning foundations - algorithms, preprocessing, feature engineering, and evaluation
Memory systems specialist for hierarchical memory, consolidation, and outcome-based learningUse when "memory system, memory hierarchy, memory consolidation, forgetting strategy,…
Use for any task that affects ML model pipeline — from raw data to live execution.
Compare model candidates using weighted metrics and deterministic ranking outputs. Use for benchmark leaderboards and model promotion decisions.
Explain ML model predictions using SHAP values, feature importance, and decision paths with visualizations.
Automated pipeline for retraining ML models with new construction data. Monitor model drift, trigger retraining, and validate model performance.
Train ML models with scikit-learn, PyTorch, TensorFlow. Use for classification/regression, neural networks, hyperparameter tuning, or encountering overfitting, underfitti — from…
Train ML models with scikit-learn, PyTorch, TensorFlow. Use for classification/regression, neural networks, hyperparameter tuning, or encountering overfitting, underfitti — from…
ML pipeline design with Metaflow, Kubeflow, and ZenML including GPU steps, artifact tracking, and production patterns.
Use when checking ML pipeline health, Metaflow flow status, failed runs, or generating an ML pipeline report.
Frame ML problems around a decision, a measurable target, and honest error costs before touching models. Use when scoping an ML project or reviewing whether one should exist.
Master ml protein folding prediction with comprehensive coverage of concepts, implementation, optimization, and production best practices.
Enforces baseline comparisons, cross-validation, interpretation, and leakage prevention for ML pipelines
Choose an ML model or provider when a user asks which checkpoint, router, local model, adapter, tokenizer, or inference backend fits a task; compare revisions, license, context,…
When the user wants to apply machine learning to supply chain problems, build ML models, or use AI for predictions.
Apply appropriate type hints for ML/PyTorch code. Use when adding type annotations to ML code or addressing mypy errors.
MLflow is an open-source platform for managing the machine learning lifecycle, including tracking, packaging, and deploy
Govern ML model lifecycle — drift detection, retraining triggers, promotion decisions
Mlflow Tracking Setup - Auto-activating skill for ML Training. Triggers on: mlflow tracking setup, mlflow tracking setup Part of the ML Training skill category.
Design or audit a model-agnostic evaluation harness for an LLM or multimodal LLM on a clinical task (radiology report generation, visual question answering, clinical text…
Master MLOps fundamentals - lifecycle, principles, tools, practices, and organizational adoption
MLPerf inference benchmarking patterns. Loadgen scenarios (SingleStream/MultiStream/Server/Offline), SUT/QSL interfaces, latency p99 targets, and performance metric…
Use when packaging an accepted MLSys paper's code, configs, and measurement scripts for the venue's post-acceptance artifact evaluation, targeting the Availability, Functional,…
Use when hardening the reproducibility of MLSys performance claims, pinning the full system layer from driver to interconnect, separating ML randomness from systems noise,…
Use when reasoning about how MLSys peer review works, covering the OpenReview workflow, the mixed ML-and-systems reviewer pool and how each half scores differently, the compressed…
Use when planning an MLSys submission cycle end to end, from venue-fit and track choice through evaluation freeze, the October deadline, the four-day January response,…
Multi-LLM intelligent switching. Use command 'multi llm' to activate local model selection based on task type. Default uses Claude Opus 4.5. — from satoshistackalotto/skills
MacBook MLX local LLM orchestration skill for running lightweight on-device models via command line. Use when you need local inference, private/offline generation, model routing…
MLX on Apple Silicon with JAX-style SplitMix64 PRNG. Deterministic color generation with GPU acceleration.
Review changes in the MLX submodule for concrete bugs, behavioral regressions, structural mismatches, and obvious performance problems.
Speech-To-Text with MLX (Apple Silicon) and opensource models (default GLM-ASR-Nano-2512) locally.
Text-To-Speech with MLX (Apple Silicon) and opensource models (default QWen3-TTS) locally.
Гейт перед /clear — read-only проверка «всё ли записано», вердикт из exit code скрипта, а не из реплики модели.
Самопроверка mm-системы — junction'ы, конфиг, vault, паспорта, bridge-архив, GSD-консистентность (passport vs PROJECT.md), наличие сторонних плагинов (karpathy, context-mode).
Simple text-to-speech skill using MiniMax Voice API. Converts text to audio with customizable voice selection. Use for generating speech audio from text.
Stage-scoped reload файлов текущего GSD-этапа в свежий контекст ПОСЛЕ ручного /clear, mid-phase. Не сводка «где мы» (это /mm resume), а ЗАГРУЗКА самих рабочих файлов этапа (STATE…
Генерирует handoff.md — компактную сводку для нового чата claude.ai когда контекст текущего заполнился.
Генерирует текст для claude.ai → Project → Instructions из шаблона + персонализирует под текущий проект. Кладёт результат в Obsidian для копипаста.
Autonomous pipeline executor — runs ALL stages from start to finish without stopping
Обзор всех зарегистрированных mm-проектов одним экраном — статус, тип, GSD-фаза, последняя активность, число открытых вопросов.
Закрывает текущую Claude Code сессию — сохраняет лог в Obsidian/Claude/Sessions/, обновляет project note, обновляет INDEX.md, перегенерирует handoff.md (для Project Knowledge…
Онбординг/персонализация mm-системы под конкретного пользователя — спрашивает имя, чем занимается, стек, язык, путь к Obsidian vault, и записывает это в личный gitignored-оверлей…
Interprets chart patterns, momentum signals, and technical setup for the discussion stage
Самообновление mm-системы из удалённого репозитория. mm раздаётся через git clone + register-skills (не npm), поэтому update = git pull свежего mm из origin/main + повторный…