目錄表

安裝 OpenCode (Ubuntu 24.04)

安裝 OpenCode (Ubuntu 24.04)

也可以參考我整合好的專案
https://github.com/tryweb/Codeforge
用 docker compose 就可啟動使用

安裝程序

curl -fsSL https://opencode.ai/install | bash

之後如果要更新也是執行

curl -fsSL https://opencode.ai/install | bash

看畫面

jonathan@claude-code-194:~$ opencode --version
1.1.19
jonathan@claude-code-194:~$ curl -fsSL https://opencode.ai/install | bash
Installed version: 1.1.19.

Installing opencode version: 1.1.21
■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■ 100%

                                 ▄
█▀▀█ █▀▀█ █▀▀█ █▀▀▄ █▀▀▀ █▀▀█ █▀▀█ █▀▀█
█░░█ █░░█ █▀▀▀ █░░█ █░░░ █░░█ █░░█ █▀▀▀
▀▀▀▀ █▀▀▀ ▀▀▀▀ ▀  ▀ ▀▀▀▀ ▀▀▀▀ ▀▀▀▀ ▀▀▀▀


OpenCode includes free models, to start:

cd <project>  # Open directory
opencode      # Run command

For more information visit https://opencode.ai/docs

預設啟動 opencode 就會自動檢查與更新至最新版本, 如果想固定在特定版本 Exp. 1.3.2

在 ~/.config/opencode/opencode.json 內加上 “autoupdate”: false

{
  "$schema": "https://opencode.ai/config.json",
  "autoupdate": false
}

安裝特定版本執行檔 Exp. 1.3.2

curl -fsSL https://opencode.ai/install | bash -s -- --version 1.3.2

使用方式

進入專案目錄執行
```
opencode
```

安裝與選用 plugin / MCP / Skills

安裝 Oh My OpenAgent

Ref - https://github.com/code-yeongyu/oh-my-openagent

sudo snap remove bun-js
sudo apt install unzip
curl -fsSL https://bun.sh/install | bash
source ~/.bashrc
cp ~/.config/opencode/opencode.json ~/.config/opencode/opencode.json.bak
bunx oh-my-openagent install

Oh My OpenAgent 內建配置熱門的 MCP Exp. context7(動態文檔), grep_app, websearch , 因此就不需要安裝這三個 MCP
Serena MCP : OpenCode 原生就支援 LSP 因此就不需要安裝 Serena MCP
ulw : UltraWork Mode → 讓所有 Agent 可以儘可能最大極限合作執行 Exp.
```
ulw 請優化後台 UI/UX 讓操作更簡易
```

ralph-loop : 自主持續迭代 → 讓 AI 持續不停優化 Exp. 搭配 ulw 使用

ulw /ralph-loop "優化首頁載入時間 <1.5s（Lighthouse 分數 >95）" --max-iterations 25 --completion-promise "Lighthouse 性能分數 >95"

mcp_tool 執行 playwrite 出現 JSON 異常問題 - 參考解法 - https://github.com/code-yeongyu/oh-my-opencode/issues/729

安裝 Claude 官方提供的 Skills

Ref - https://github.com/anthropics/skills

安裝 ui-ux-pro-max-skill

Ref - https://github.com/nextlevelbuilder/ui-ux-pro-max-skill

# Install CLI globally
sudo npm install -g uipro-cli

# Go to your project
cd /path/to/your/project
# Install for your AI assistant
uipro init --ai opencode    # OpenCode

重新啟動 opencode 輸入 “Build a landing page for a SaaS product”

安裝 openspec

Ref - https://github.com/Fission-AI/OpenSpec

Step 1.

sudo npm install -g @fission-ai/openspec@latest
jonathan@claude-code-194:~$ openspec --version
0.21.0

Step 2.

cd my-project
openspec init

看畫面

 ████   █████   ██████  ██  ██   █████  █████   ██████   █████
██  ██  ██  ██  ██      ███ ██  ██      ██  ██  ██      ██
██  ██  █████   █████   ██ ███   ████   █████   █████   ██
██  ██  ██      ██      ██  ██      ██  ██      ██      ██
 ████   ██      ██████  ██  ██  █████   ██      ██████   █████

Welcome to OpenSpec!

Step 3/3

Review selections
Press Enter to confirm or Backspace to adjust.

▌ Antigravity
▌ OpenCode
▌ OpenSpec structure created
▌ AI tools configured

✔ OpenSpec initialized successfully!

Tool summary:
▌ Root AGENTS.md stub refreshed for other assistants
▌ Created: Antigravity, and OpenCode
▌ Skipped: Amazon Q Developer, Auggie, Claude Code, Cline, Codex, CodeBuddy Code, Continue (VS Code / JetBrains / Cli), CoStrict, Crush, Cursor, Factory Droid, Gemini CLI, GitHub Copilot, iFlow, Kilo Code, Qoder, Qwen Code, RooCode, and Windsurf

Use `openspec update` to refresh shared OpenSpec instructions in the future.

Important: Restart your IDE
Slash commands are loaded at startup. Please restart your coding assistant
to ensure the new /openspec commands appear in your command palette.

Next steps - Copy these prompts to Antigravity, and OpenCode:
────────────────────────────────────────────────────────────
1. Populate your project context:
   "Please read openspec/project.md and help me fill it out
    with details about my project, tech stack, and conventions"

2. Create your first change proposal:
   "I want to add [YOUR FEATURE HERE]. Please create an
    OpenSpec change proposal for this feature"

3. Learn the OpenSpec workflow:
   "Please explain the OpenSpec workflow from openspec/AGENTS.md
    and how I should work with you on this project"
────────────────────────────────────────────────────────────

複製畫面看到的這段

1. Populate your project context:
   "Please read openspec/project.md and help me fill it out
    with details about my project, tech stack, and conventions"

2. Create your first change proposal:
   "I want to add [YOUR FEATURE HERE]. Please create an
    OpenSpec change proposal for this feature"

3. Learn the OpenSpec workflow:
   "Please explain the OpenSpec workflow from openspec/AGENTS.md
    and how I should work with you on this project"

等等貼到 opencode , 分析目前專案狀態產生 openspec/project.md (PRD 檔), 當然也可以使用原本的 PRD 檔案複製成 project.md

進入 opencode 執行 /openspec 命令

Draft the Proposal Exp.

/openspec:proposal Add profile search filters

Verify & Review

$ openspec list                             # Confirm the change folder exists
$ openspec validate add-profile-filters     # Validate spec formatting
$ openspec show add-profile-filters         # Review proposal, tasks, and spec delta

Refine the Specs

請修正 xxx 功能, 將 abc 改成 zyx

Implement the Change

The specs look good. Let's implement this change.

Archive the Completed Change

$ openspec archive add-profile-filters --yes  # Archive the completed change without prompts

查看規格清單
```
openspec list
```

更版方式

sudo npm install -g @fission-ai/openspec@latest
cd myproject
openspec update

使用 openchamber Web 操作介面

設定 openchamber 可以在 Ubuntu 開機自動啟動

參考網址 - - https://github.com/btriapitsyn/openchamber

# Quick install (auto-detects your package manager)
curl -fsSL https://raw.githubusercontent.com/btriapitsyn/openchamber/main/scripts/install.sh | bash

# Quick start daemon
openchamber --port 6001 --daemon

# 其他參數
openchamber                          # Start on port 3000
openchamber --port 8080              # Custom port
openchamber --daemon                 # Background mode
openchamber --ui-password secret     # Password-protect UI
openchamber --try-cf-tunnel          # Create a Cloudflare Quick Tunnel for remote access
openchamber --try-cf-tunnel --tunnel-qr              # Show QR code for easy mobile access
openchamber --try-cf-tunnel --tunnel-password-url   # Include password in URL for auto-login
OPENCODE_PORT=4096 OPENCODE_SKIP_START=true openchamber  # Connect to external OpenCode server
openchamber stop                     # Stop server
openchamber update                   # Update to latest version

參考網址

ai coding