Copy this prompt into the AI assistant you're using.
Follow https://skillsmp.com/skill-install/prompt.md to install Agent Skills from https://github.com/ScaleDockHQ/llm-skills.Loading the repository README…
Showing 40 of 215 collected skills.
Install and operate AIBrix v0.7.0 on Kubernetes, the vLLM project's control plane for LLM serving, with an Envoy gateway and LLM-aware routing, PodAutoscaler, LoRA ModelAdapter, StormService prefill/decode and multi-node, and KVCache offload. Use for…
Aider, the terminal AI pair-programming tool (aider-chat, `aider`), pointed at self-hosted or gateway models through LiteLLM-style model names. Use to install or upgrade aider, set OPENAI_API_BASE with the `openai/` prefix, `ollama_chat/` or…
Set up, verify and operate AMD Instinct GPUs (MI300X, MI325X, MI350X, MI355X, MI250, MI210) with ROCm 10.x for LLM inference. Use for amdgpu driver and ROCm install, gfx942/gfx950 targets, /dev/kfd and render group errors, amd-smi (rocm-smi removed),…
Set up, size and operate AMD Ryzen AI Max (Strix Halo, gfx1151, Radeon 8060S, up to 128 GB unified memory) for local LLM inference on Linux or Windows. Use for GTT/TTM unified-memory limits (ttm.pages_limit, amdgpu.gttsize, amd-ttm), BIOS UMA carveout, Vulkan…
Use the Claude API directly (api.anthropic.com) - keys, workspaces, Messages API, OpenAI SDK compatibility, streaming, tools, structured outputs, Message Batches, prompt caching, Files, rate-limit tiers, spend limits, ZDR, key revocation. Use for…
Anthropic Python and TypeScript SDKs pointed at self-hosted endpoints. Use to set base_url/baseURL, ANTHROPIC_AUTH_TOKEN vs ANTHROPIC_API_KEY for vLLM, llama.cpp, Ollama, LiteLLM or llama-swap Messages routes, map model names, stream, call tools, or fix 404…
AnythingLLM, all-in-one chat and RAG app with workspaces, agents and MCP (Desktop or Docker). Use to point it at vLLM, llama.cpp, Ollama, LiteLLM or llama-swap with the Generic OpenAI, Ollama or LiteLLM provider, keep keys in the server .env, set context…
Install, serve and tune Aphrodite Engine (now branded Sonar), the vLLM-derived LLM inference server with OpenAI, Anthropic and Kobold APIs. Use for the aphrodite CLI, aphrodite serve, aphrodite-engine pip package, sonar.dphn.ai, port 2242, quantization (exl3,…
Size, tune and operate Apple Silicon Macs (M-series MacBook, Mac mini, Mac Studio) for local LLM inference with Metal and MLX. Use for unified memory sizing, iogpu.wired_limit_mb, "[WARNING] Generating with a model that requires",…
Arize Phoenix, open-source LLM tracing, evals and prompt playground (arize-phoenix, arizephoenix/phoenix, phoenix serve). Use to self-host it (Docker, Postgres, Helm), send OpenInference or OTLP traces from apps, LiteLLM and vLLM, set auth, retention and…
Use Amazon Bedrock to call hosted LLMs from your AWS account via Converse, InvokeModel, OpenAI-compatible Chat Completions or Responses, and Anthropic Messages. Covers model access, inference profiles and cross-Region inference, API keys, IAM, VPC endpoints,…
Set up and operate AWS Inferentia2 and Trainium (Inf2, Trn1, Trn2, Trn3) with the Neuron SDK to serve LLMs through vLLM Neuron or NxD Inference, using Neuron DLAMIs and DLCs. Use for neuron-ls, vllm-neuron, NxD Inference, NEURON_LOGICAL_NC_CONFIG,…
Host your own LLM on Amazon SageMaker AI endpoints (real-time, async, serverless, inference components), with JumpStart and LMI/vLLM containers, scale to zero, SigV4 invocation, teardown and cost checks. Use for SageMaker, JumpStart, LMI container,…
Axolotl is a YAML-driven LLM fine-tuning framework (axolotl CLI) for LoRA, QLoRA, full fine-tuning, SFT, DPO, KTO and GRPO on Hugging Face models. Use to install it (uv, axolotlai/axolotl Docker, arm64), write and debug configs, run axolotl…
Use Microsoft Foundry, formerly Azure AI Foundry and Azure OpenAI, to deploy and call GPT, Claude and open models with the az CLI, deployment types, the v1 API, Entra ID auth, quotas, budgets, Azure ML online endpoints and teardown. Use for Azure OpenAI,…
Baseten model APIs and dedicated GPU deployments - call hosted models over OpenAI or Anthropic-style endpoints, or deploy your own with the baseten CLI and Truss (config.yaml, vLLM or Engine-Builder), scale, monitor cost and tear down. Use for "Baseten",…
Install and run Bifrost (maximhq/bifrost), a Go open-source AI gateway with a web UI that fronts 20+ providers behind OpenAI-, Anthropic-, Bedrock- and GenAI-compatible APIs, with fallbacks, virtual keys, budgets, semantic cache and MCP. Use when the user…
Load and train LLMs in 8-bit and 4-bit with bitsandbytes (LLM.int8(), NF4/FP4 QLoRA, 8-bit and paged optimizers) through Transformers, PEFT, TRL and Accelerate. Use for BitsAndBytesConfig, load_in_4bit, QLoRA or FSDP-QLoRA setup, Adam8bit and paged_adamw,…
Caddy is a web server and reverse proxy with automatic HTTPS, used as a loopback front for LLM endpoints. Use to install it, write a Caddyfile that proxies to LiteLLM, vLLM or Ollama with SSE streaming and long timeouts, add bearer or forward_auth checks,…
Use Cerebras Inference Cloud (api.cerebras.ai) for fast OpenAI-compatible open-model APIs, with key and project setup, model listing, rate limits, prompt caching, batch, dedicated endpoints, LiteLLM and teardown. Use when the user mentions Cerebras,…
Install and run Chatterbox, Resemble AI's MIT-licensed open text-to-speech family (Turbo 350M, Nano 110M, Multilingual V3 with 23 languages, original English) with zero-shot voice cloning, [laugh] paralinguistic tags, exaggeration and cfg_weight, voice…
Choose and size hardware for running an LLM, from model size, quantization, context and concurrency, using VRAM and memory-bandwidth math plus a spec table for NVIDIA RTX, datacenter, Blackwell and DGX Spark, AMD Instinct, Radeon and Strix Halo, Apple Silicon…
Choose an LLM inference engine from hardware, model format, features and concurrency, with a decision flow and an engine matrix for vLLM, SGLang, TensorRT-LLM, llama.cpp, Ollama, TensorFold and ExLlamaV3. Use when asked which engine to run, vLLM vs SGLang vs…
Decide where LLMs should run and record the decision. Compares seven hosting tiers (laptop or desktop local, owned server such as a DGX Spark or GPU box, rented GPU VM, serverless GPU, managed open-model API, frontier proprietary API, hyperscaler platform…
Run Chroma, the simple open-source vector database, beside a self-hosted LLM stack for RAG. Use for chromadb/chroma Docker, port 8000, the HttpClient, OpenAI-compatible embedding functions pointed at LiteLLM or TEI, RAM sizing (HNSW lives in memory), fronting…
Claude Code, Anthropic's terminal coding agent, pointed at self-hosted or gateway models over the Anthropic Messages API. Use to install or update `claude`, set ANTHROPIC_BASE_URL, ANTHROPIC_AUTH_TOKEN or apiKeyHelper for vLLM, llama.cpp, Ollama, LiteLLM or…
Configure Cline (VS Code, JetBrains, CLI, desktop) for self-hosted LLMs with the OpenAI Compatible, Anthropic custom base URL, Ollama or LiteLLM providers (base URL, key handling, model ID, context window and output limits, compact prompt, proxies). Use for…
Run and consume LLMs in the cloud - pick a tier, provision, deploy, connect to one gateway endpoint, secure, set budgets, benchmark, scale, update and tear down. Covers frontier APIs (OpenAI, Anthropic Claude, Google Gemini), managed open-model APIs (Together…
Operate Cloudflare AI Gateway, a managed proxy that adds logging, caching, rate and spend limits, retries and dynamic routing in front of OpenAI, Anthropic, Workers AI and other providers. Use when the user mentions AI Gateway, gateway.ai.cloudflare.com,…
Cloudflare Tunnel (cloudflared) publishes a local LLM endpoint at a public HTTPS hostname with no open port, optionally behind Cloudflare Access. Use to install cloudflared, write ingress rules, run it under systemd, put a dashboard behind Access, give API…
OpenAI Codex CLI, the terminal coding agent, pointed at self-hosted or gateway models through custom model providers over the Responses API. Use to install or update `codex`, write ~/.codex/config.toml model_providers (base_url, env_key, wire_api responses)…
Install, run headless and operate ComfyUI, the node-graph engine for image, video and audio diffusion models (FLUX, SDXL, Wan, LTX), with its HTTP API, custom nodes and model folders. Use for python main.py --listen/--port, workflow_api.json, /prompt and…
Configure Continue (VS Code, JetBrains, cn CLI) for self-hosted LLMs via ~/.continue/config.yaml, with apiBase, secrets in ~/.continue/.env, roles, capabilities and model-name mapping for vLLM, llama.cpp, Ollama, LiteLLM and llama-swap, OpenAI- or…
Run LLMs on CoreWeave, CKS managed Kubernetes GPU clusters and Node Pools, SUNK (Slurm on Kubernetes), Serverless Inference (api.inference.wandb.ai, W&B/Forge) and Dedicated Inference (gateways and deployments via the cwic CLI). Use for "CoreWeave", "CKS",…
Run, size and tune LLM inference on CPUs only (x86 AVX2/AVX-512/AMX, AArch64 servers with NEON and KleidiAI, Intel Macs) with llama.cpp. Use for no-GPU servers, VMs, "llama-server is slow on CPU", threads and NUMA (--numa, -t, --cpu-mask), GGML_NATIVE,…
Run LLMs on Crusoe Cloud, GPU VMs and Managed Kubernetes (CMK) through the crusoe CLI, plus Crusoe Managed Inference (Intelligence Foundry) with an OpenAI-compatible serverless API and self-serve dedicated deployments. Use for "Crusoe", "crusoe compute vms…
Connect Cursor to your own OpenAI-compatible LLM endpoint (vLLM, llama.cpp, Ollama, LiteLLM, llama-swap) with Settings > Models, Override OpenAI Base URL, custom model names and a scoped key. Use for Cursor BYOK, "local model in Cursor", localhost or…
Datadog Agent Observability (formerly LLM Observability): traces, cost, evaluations and ml_obs metrics for LLM apps. Use to instrument Python or Node with ddtrace and dd-trace, trace vLLM, llama.cpp, SGLang, Ollama or a LiteLLM proxy, send OTLP, redact with…
NVIDIA DCGM Exporter, the Prometheus exporter for GPU telemetry (utilization, memory, power, temperature, XID errors, profiling counters). Use to install it as a package, Docker container or Helm release, pick collector CSV fields, scrape it from Prometheus,…
Run DeepEval (`deepeval test run`, pytest-based LLM unit tests) with GEval, answer relevancy, faithfulness and other metrics, using a local or OpenAI-compatible judge (vLLM, Ollama, LiteLLM) instead of the OpenAI default. Use for set-local-model, telemetry…
The repository README could not be loaded. You can still explore its Skills below.