AgentSkillsCN

logics-ollama-specialist

在本地应用中安装、配置并集成Ollama（如macOS/Homebrew，或通过OLLAMA_HOST/OLLAMA_ORIGINS等环境变量），并通过Vite或Express代理将前端与后端连接至/api/chat。当需要构建或调试基于Ollama的UI界面、解决局域网内的CORS/403 Origin问题，或搭建调用Ollama的PWA/本地开发流程时，这一工具便能大显身手。

SKILL.md
--- frontmatter
name: logics-ollama-specialist
description: Install, configure, and integrate Ollama in local apps (macOS/Homebrew, env vars like OLLAMA_HOST/OLLAMA_ORIGINS) and wire frontend/backends to /api/chat via Vite or Express proxies. Use when building or debugging Ollama-powered UIs, fixing CORS/403 Origin issues on LAN, or setting up PWA/local dev flows that call Ollama.

Logics Ollama Specialist

Quick start

  1. •Clarify OS, target model, and whether Ollama should run locally or on another host.
  2. •If installation is needed, ask for explicit permission before running install commands.
  3. •If the app is a frontend, prefer a same-origin proxy (/ollama/*) to avoid CORS.
  4. •If LAN URL causes 403, remove Origin in the proxy or set OLLAMA_ORIGINS.

Install Ollama (macOS)

  • •Ask for explicit approval before running installs.
  • •Use the script for safe/consistent setup:
bash
DRY_RUN=1 logics/skills/logics-ollama-specialist/scripts/ollama_install_macos.sh qwen3

If approved, run again without DRY_RUN=1.

Install and manage CodeLlama

Use any CodeLlama tag (codellama, codellama:7b, codellama:13b, codellama:34b).

  • •Pull/install a model:
bash
ollama pull codellama:7b
  • •Or with the install script:
bash
DRY_RUN=1 logics/skills/logics-ollama-specialist/scripts/ollama_install_macos.sh codellama:7b
  • •List installed CodeLlama models:
bash
ollama list | grep -E '^codellama'
  • •Run and test quickly:
bash
ollama run codellama:7b
  • •Remove a local model:
bash
ollama rm codellama:7b

Start and stop Ollama server

  • •Start in foreground:
bash
ollama serve
  • •Start as Homebrew service:
bash
brew services start ollama
  • •Stop foreground server: Ctrl + C
  • •Stop Homebrew service:
bash
brew services stop ollama
  • •If Ollama.app auto-restarts the server, quit the app:
bash
osascript -e 'quit app "Ollama"'
  • •Last-resort force stop:
bash
pkill -f 'ollama serve'
pkill -x Ollama
  • •Verify server status:
bash
lsof -iTCP:11434 -sTCP:LISTEN -n -P

Verify Ollama

Run the check script to validate prerequisites and reachability:

bash
logics/skills/logics-ollama-specialist/scripts/ollama_check.sh qwen3

Integrate with a frontend (Vite/React)

  1. •Use a proxy to call /ollama/* from the browser.
  2. •Set VITE_OLLAMA_HOST to the local Ollama URL.
  3. •If LAN access returns 403, remove the Origin header in the proxy or set OLLAMA_ORIGINS.

See references/ollama-integration.md for snippets and common pitfalls.

Add PWA support (Vite)

Use vite-plugin-pwa, register the SW, and add icons/manifest fields. See references/pwa-vite.md.

Resources

scripts/

  • •ollama_check.sh: Verify ollama, Node, and basic /api/version reachability.
  • •ollama_install_macos.sh: Install/start Ollama with an optional model (supports DRY_RUN=1).

references/

  • •ollama-integration.md: Proxy/CORS/Origin fixes, env vars, and endpoint notes.
  • •pwa-vite.md: PWA steps for Vite/React apps.