菜单
Model server administrator for managing OpenCode providers, GPU optimization, and local LLM endpoints. - opencode - local_models - model_management - install_models - check_hardware - "@mcp-manager" - "@gemini-conductor"
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Professional bootstrapper for Vercel-style Agentic Repositories using compressed AGENTS.md manifests.
Persistent SQLite-based cognitive layer for cross-session interaction logging and insight distillation.
Design lead for C4 structural documentation, Mermaid.js diagrams, and Architecture Decision Records (ADRs).
Specialist in core AWS infrastructure: EC2 compute, IAM identity, VPC networking, and RDS/Aurora databases.
5-phase migration engine for transitioning workloads from Google Cloud Platform to AWS architecture.
AI/ML infrastructure expert for SageMaker fine-tuning, HyperPod cluster management, and dataset evaluation.
| name | skill-opencode |
| description | Model server administrator for managing OpenCode providers, GPU optimization, and local LLM endpoints. - opencode - local_models - model_management - install_models - check_hardware - "@mcp-manager" - "@gemini-conductor" |
Expert in managing the opencode platform for local LLM execution. Optimized for NVIDIA RTX 3060 (6GB VRAM).
check-hardwareDetects GPU/CPU/RAM and recommends the best SOTA models for the current tier.
python3 {PROJECT_DIR}/bin/model_selector.py checkinstall-best-modelsDownloads and installs the GPU-optimized 2026 model stack.
python3 {PROJECT_DIR}/bin/model_selector.py installopencode serve (Starts headless server)opencode web (Opens browser TUI)opencode stats (Token/Cost usage)| Model | Usage | Target Device |
|---|---|---|
| Qwen 2.5 Coder 7B | Primary Agentic Work (Tools/Coding) | GPU (Fits in 6GB) |
| DeepSeek R1 8B | Complex Reasoning / Math / Logic | GPU (Fits in 6GB) |
| Gemma 4 E4B | Ultra-fast Edge Reasoning | GPU (Instant) |
| Llama 3.2 3B | Lightweight Chat Fallback | GPU/CPU |
Local execution on this system is currently configured for GPU Acceleration via CUDA 12.7. Expect speeds of 30-50 tokens/second on primary models.