with one click
IntelliSoins-agent
IntelliSoins-agent contains 23 collected skills from IntelliSoins, with repository-level occupation coverage and site-owned skill detail pages.
Skills in this repository
Proton CLI local email indexing and search pipeline.
Proton CLI local email indexing and search pipeline.
Expert LiteLLM - AI Gateway proxy multi-provider, routing fallback, budgets/spend, caching, guardrails, MCP gateway, logging. Consulter pour decisions d'integration LiteLLM dans IntelliSoins.
mlx-audio — Text-to-Speech (TTS), Speech-to-Text (STT) et Speech-to-Speech (S2S) optimisé pour Apple Silicon. Supporte le clonage de voix zero-shot (VoxCPM2, F5-TTS, Voxtral) et le fine-tuning (OuteTTS via mlx-lm, F5-TTS).
Manage the local AI/ML inference servers in this repository via `./aictl`. Use this project skill whenever Michael asks about ai-servers, aictl, local model server status, health checks, LaunchAgents, port conflicts, MLX servers, LiteLLM proxy, embeddings, reranker, GLiNER, Whisper STT, Kokoro TTS, oMLX, Ollama, or the PRO-G40 model-cache disk.
Stack AI/ML personnelle Michael Ahern, installée nativement sur le disque (Apple Silicon M3 Max via Homebrew + Launch...
Clustering des pain points pharmacie : Discovery (problème de burnout, scraping PharmQC) → Development (NLP français,...
mlx-omni-server (madroidmaq/mlx-omni-server) — serveur d'inférence local Apple Silicon, dual API OpenAI (/v1/*) ET Anthropic (/anthropic/v1/*) sur le même port, auto-discovery zero-config depuis le cache HuggingFace, suite complète (chat tools/streaming/structured output, audio TTS+STT, image-gen, embeddings), function calling, thinking mode. Charge on-demand sur fichiers mlx-omni-server ou servers.yaml.
mlx-openai-server (cubist38/mlx-openai-server) — serveur OpenAI-compatible Apple Silicon, 6 types de modèles (lm, multimodal, image-gen, image-edit, embeddings, whisper), config YAML multi-modèle on-demand, multi-adapter LoRA (--lora-paths), 11+ parsers tool/reasoning, KV cache quantization (q4/q8), speculative decoding, image-gen/edit via mflux, Whisper, structured output. Charge on-demand sur fichiers mlx-openai-server ou servers.yaml.
Carte de la stack MLX locale (Apple Silicon) — frameworks d'inférence/lib Apple, 8 serveurs d'inférence, fine-tuning, optimisation KV. Index qui pointe vers les rules détaillées + skills intellisoins-mlx, versions upstream au 2026-05-29. Charge on-demand sur fichiers MLX, servers.yaml ou stack ai-servers.
Operate, test, and diagnose Michael's local medical NER, RAG, pgvector, KG, BGE reranking, and Apache AGE pipeline. Use this project skill whenever the user mentions NER-RAG-KG-reranking, kg_pipeline, GLiNER biomedical NER, pgvector embeddings, BGE reranker, SNOMED KG, Apache AGE medical_graph, edge candidates, ontology validation, or asks to test a clinical query against the local `medical` database.
oMLX (jundot/omlx) — serveur d'inférence MLX local Apple Silicon : continuous batching, tiered KV cache (RAM chaud + SSD froid persistant aux restarts), backend Claude Code/Codex, multi-modèle (LRU/pinning/TTL), VLM+OCR, embeddings+rerankers, API OpenAI+Anthropic native, menubar macOS. Charge on-demand sur fichiers omlx ou servers.yaml.
Use this project skill when Michael asks to configure OpenClaw providers, add a local LLM, switch model, configure Qwen3.5, inspect OpenClaw tools, subagents, system prompts, Pi SDK agent runtime, or model provider config.
Use this project skill when Michael asks to build OpenClaw, run tests, fix build errors, run pnpm install, lint, format, merge upstream, or troubleshoot OpenClaw/OpenIntellisoins development failures.
Use this project skill when Michael asks to fine-tune for OpenClaw, run the pipeline, build training datasets, run GRPO, create LoRA adapters, extract Claude sessions, convert Anthropic to OpenAI format, enrich OpenClaw tools, or work in the OpenClaw pipeline directory.
Use this project skill when Michael asks to build the OpenClaw UI, translate OpenClaw, add i18n keys, fix the Control UI, work with Lit components, change the dashboard header, or modify the `ui/` web interface.
Use this project skill when Michael asks to start the OpenClaw gateway, configure OpenClaw, fix gateway hangs, compare gateway dev vs prod, edit openclaw.json, troubleshoot WebSocket connections, auth, or gateway health.
Catalogue des datasets d'entrainement pour fine-tuning local sur Apple Silicon (M3 Max 128 GB).
TurboQuant — compression KV-cache pour inférence MLX sur Apple Silicon (3-5× via rotation PolarQuant + quantization, Google Research mars 2026). 6 chemins d'intégration (mlx-vlm, mlx-openai-server, mlx-optiq, llama.cpp, SwiftLM, helgklaizar v1). Patch MoE-aware Qwen3.5. Déployé dans ~/ai-servers/. Charge on-demand sur fichiers turboquant ou servers.yaml.
Official community-maintained hardware plugin enabling vLLM inference on Apple Silicon. Uses MLX as primary compute b...
vLLM-MLX (waybarrios) — serveur d'inférence OpenAI + Anthropic compatible sur Apple Silicon. LLM/VLM avec continuous batching, MCP tool calling, multimodal, /v1/rerank (BERT) et embeddings. 400+ tok/s natif MLX. Supporte Qwen3/3.5, Llama, Gemma 4, DeepSeek-R1. Charge on-demand sur fichiers vllm-mlx ou servers.yaml.
vLLM-Omni — serving de modèles omni-modalité (texte, image, vidéo, audio) sur GPU/NPU CUDA/ROCm via API OpenAI unifiée, étend vLLM aux Diffusion Transformers (DiT). NON supporté sur Apple Silicon (utiliser mflux/mlx-video/mlx-audio). Charge on-demand sur fichiers vllm-omni.
Moteur d'inférence MLX local sur Apple Silicon (vMLX, Jinho Jang) — LLM/vision/image-gen/audio/multimodal Nemotron-Omni, 5-layer KV caching, JANG quantization, inférence distribuée multi-Mac, API Anthropic+OpenAI+Ollama native. Charge on-demand quand on travaille dans la stack ai-servers ou sur des fichiers vmlx.