القائمة
Use for llama.cpp, llama-server, local model hosting, ROCm or Vulkan inference, model and KV-cache tuning, Open WebUI, or OpenCode local-provider configuration in this repository.
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
Use for diagnosing host freezes, GPU or kernel instability, hardware-specific failures, boot issues, and prior goldenball freeze investigations in this repository.
Use for creating, modifying, reviewing, or debugging Incus/LXC container images, profiles, cloud-init, networking, GPU passthrough, deployment, or runtime behavior in this repository.
Use for Neovim, nixvim, editor plugins, keymaps, LSP, completion, themes inside Neovim, or OpenCode's Neovim integration in this repository.
Use for shared colors, palettes, oxocarbon styling, terminal or application themes, and keeping theme values consistent across this repository.
استنادا إلى تصنيف SOC المهني
| name | llm-hosting |
| description | Use for llama.cpp, llama-server, local model hosting, ROCm or Vulkan inference, model and KV-cache tuning, Open WebUI, or OpenCode local-provider configuration in this repository. |
Read the relevant section of docs/LLM-HOSTING-TUNING.md; search its headings first instead of loading unrelated sections.
For OpenCode and nixvim integration, also read docs/NIXVIM.md and inspect modules/home/nixvim/default.nix. For NixOS service behavior, inspect modules/nixos/llama-cpp/default.nix.
Preserve measured hardware-specific values unless the task provides new evidence. After edits, run ./format.fish; never run a Nix build or activation command.
