From 1324f1575d2bad2b02134996a84788def04b6991 Mon Sep 17 00:00:00 2001
From: Jeffery <jiantw83@yahoo.com>
Date: Mon, 11 May 2026 07:23:06 +0000
Subject: [PATCH 1/7] =?UTF-8?q?feat:=20=E9=9A=8E=E6=AE=B5=E4=B8=80=20-=20?=
 =?UTF-8?q?=E5=9F=BA=E6=9C=AC=E6=B5=81=E7=A8=8B=E4=B8=B2=E6=8E=A5=E9=AA=A8?=
 =?UTF-8?q?=E6=9E=B6?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

- 重寫 action.yaml：支援所有 LLM providers 的 inputs
- 重寫 Dockerfile：python:3.11-slim + git
- 重寫 entrypoint.sh：啟動 app/main.py
- app/config.py：環境變數與 LLM 自動偵測
- app/llm.py：OpenAI-compatible 統一介面
- app/gitea.py：PR diff 取得與 comment 發布
- app/roles.py：從 prompts/roles/*.yaml 載入角色
- app/main.py：pipeline 骨架，log 每個主要階段
- app/prompts/roles/：五個角色定義（Aria/Rex/Zara/Leo/Maya）
---
 Dockerfile                             |  17 ++-
 action.yaml                            | 166 ++++++++++++++++++++++---
 app/config.py                          |  31 +++++
 app/gitea.py                           |  26 ++++
 app/llm.py                             |  50 ++++++++
 app/main.py                            |  75 +++++++++++
 app/prompts/roles/maintainability.yaml |  23 ++++
 app/prompts/roles/performance.yaml     |  23 ++++
 app/prompts/roles/security.yaml        |  23 ++++
 app/prompts/roles/style.yaml           |  23 ++++
 app/prompts/roles/testing.yaml         |  23 ++++
 app/requirements.txt                   |   3 +
 app/roles.py                           |  22 ++++
 entrypoint.sh                          |  13 +-
 14 files changed, 490 insertions(+), 28 deletions(-)
 create mode 100644 app/config.py
 create mode 100644 app/gitea.py
 create mode 100644 app/llm.py
 create mode 100644 app/main.py
 create mode 100644 app/prompts/roles/maintainability.yaml
 create mode 100644 app/prompts/roles/performance.yaml
 create mode 100644 app/prompts/roles/security.yaml
 create mode 100644 app/prompts/roles/style.yaml
 create mode 100644 app/prompts/roles/testing.yaml
 create mode 100644 app/requirements.txt
 create mode 100644 app/roles.py

diff --git a/Dockerfile b/Dockerfile
index af3dacb..1ef0bb9 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -1,10 +1,15 @@
-FROM alpine:latest
+FROM python:3.11-slim
 
-# 安裝必要的工具
-RUN apk add --no-cache --no-check-certificate bash
-    
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    git \
+    && rm -rf /var/lib/apt/lists/*
+
+WORKDIR /action
+
+COPY app/ /action/app/
 COPY entrypoint.sh /entrypoint.sh
 
-RUN chmod +x /entrypoint.sh
+RUN pip install --no-cache-dir -r /action/app/requirements.txt && \
+    chmod +x /entrypoint.sh
 
-ENTRYPOINT ["/entrypoint.sh"]
\ No newline at end of file
+ENTRYPOINT ["/entrypoint.sh"]
diff --git a/action.yaml b/action.yaml
index f8849d0..4e28a5c 100644
--- a/action.yaml
+++ b/action.yaml
@@ -1,20 +1,158 @@
-name: 'Docker Action Template'
-description: 'Docker Action 範本'
+name: 'AI Code Review'
+description: 'AI 多角色 Code Review Action，自動分析 PR 並發布問題 Comment'
 author: 'Jeffery'
 inputs:
-  runner_token:
-    description: 'Gitea Runner Token'
-    required: true
-  text:
-    description: '輸入的文字'
-    default: "Hello, World!"
-outputs:
-  text:
-    description: '輸出的文字'
+  # Gitea 相關（可從 gitea context 自動取得）
+  GITEA_TOKEN:
+    description: 'Gitea API Token'
+    required: false
+  GITEA_SERVER_URL:
+    description: 'Gitea Server URL'
+    required: false
+  GITEA_REPOSITORY:
+    description: 'Gitea Repository (owner/repo)'
+    required: false
+  PR_NUMBER:
+    description: 'Pull Request Number'
+    required: false
+  PR_HEAD_BRANCH:
+    description: 'PR 來源分支'
+    required: false
+  PR_BASE_BRANCH:
+    description: 'PR 目標分支'
+    required: false
+
+  # OpenAI-compatible
+  OPENAI_API_KEY:
+    description: 'OpenAI / OpenRouter API Key'
+    required: false
+  OPENAI_BASE_URL:
+    description: 'OpenAI-compatible Base URL'
+    required: false
+  OPENAI_MODEL:
+    description: 'OpenAI-compatible Model Name'
+    required: false
+
+  # Anthropic Claude
+  CLAUDE_API_KEY:
+    description: 'Anthropic Claude API Key'
+    required: false
+  CLAUDE_BASE_URL:
+    description: 'Claude Base URL'
+    required: false
+  CLAUDE_MODEL:
+    description: 'Claude Model Name'
+    required: false
+
+  # Google Gemini
+  GEMINI_API_KEY:
+    description: 'Google Gemini API Key'
+    required: false
+  GEMINI_BASE_URL:
+    description: 'Gemini Base URL'
+    required: false
+  GEMINI_MODEL:
+    description: 'Gemini Model Name'
+    required: false
+
+  # Ollama
+  OLLAMA_BASE_URL:
+    description: 'Ollama Base URL'
+    required: false
+  OLLAMA_MODEL:
+    description: 'Ollama Model Name'
+    required: false
+
+  # Amazon Q
+  AMAZONQ_API_KEY:
+    description: 'Amazon Q API Key'
+    required: false
+  AMAZONQ_BASE_URL:
+    description: 'Amazon Q Base URL'
+    required: false
+
+  # SonarQube
+  SONARQUBE_TOKEN:
+    description: 'SonarQube Token'
+    required: false
+  SONARQUBE_URL:
+    description: 'SonarQube URL'
+    required: false
+
+  # Kilo Code
+  KILO_API_KEY:
+    description: 'Kilo Code API Key'
+    required: false
+  KILO_BASE_URL:
+    description: 'Kilo Code Base URL'
+    required: false
+
+  # Roo Code
+  ROO_API_KEY:
+    description: 'Roo Code API Key'
+    required: false
+  ROO_BASE_URL:
+    description: 'Roo Code Base URL'
+    required: false
+
+  # Cline
+  CLINE_API_KEY:
+    description: 'Cline API Key'
+    required: false
+  CLINE_BASE_URL:
+    description: 'Cline Base URL'
+    required: false
+
+  # Continue
+  CONTINUE_API_KEY:
+    description: 'Continue API Key'
+    required: false
+  CONTINUE_BASE_URL:
+    description: 'Continue Base URL'
+    required: false
+
+  # Kade
+  KADE_API_KEY:
+    description: 'Kade API Key'
+    required: false
+  KADE_BASE_URL:
+    description: 'Kade Base URL'
+    required: false
+
 runs:
   using: 'docker'
   image: 'Dockerfile'
   env:
-    GITEA_SERVER_URL: ${{ gitea.server_url }}
-    GITEA_REPOSITORY: ${{ gitea.repository }}
-    RUNNER_TOKEN: ${{ inputs.runner_token || secrets.GITEA_TOKEN || secrets.RUNNER_TOKEN }}
\ No newline at end of file
+    # Gitea context（優先用 inputs，否則從 gitea context 取）
+    GITEA_TOKEN: ${{ inputs.GITEA_TOKEN || secrets.GITEA_TOKEN }}
+    GITEA_SERVER_URL: ${{ inputs.GITEA_SERVER_URL || gitea.server_url }}
+    GITEA_REPOSITORY: ${{ inputs.GITEA_REPOSITORY || gitea.repository }}
+    PR_NUMBER: ${{ inputs.PR_NUMBER || gitea.event.pull_request.number }}
+    PR_HEAD_BRANCH: ${{ inputs.PR_HEAD_BRANCH || gitea.event.pull_request.head.ref }}
+    PR_BASE_BRANCH: ${{ inputs.PR_BASE_BRANCH || gitea.event.pull_request.base.ref }}
+    # LLM
+    OPENAI_API_KEY: ${{ inputs.OPENAI_API_KEY }}
+    OPENAI_BASE_URL: ${{ inputs.OPENAI_BASE_URL }}
+    OPENAI_MODEL: ${{ inputs.OPENAI_MODEL }}
+    CLAUDE_API_KEY: ${{ inputs.CLAUDE_API_KEY }}
+    CLAUDE_BASE_URL: ${{ inputs.CLAUDE_BASE_URL }}
+    CLAUDE_MODEL: ${{ inputs.CLAUDE_MODEL }}
+    GEMINI_API_KEY: ${{ inputs.GEMINI_API_KEY }}
+    GEMINI_BASE_URL: ${{ inputs.GEMINI_BASE_URL }}
+    GEMINI_MODEL: ${{ inputs.GEMINI_MODEL }}
+    OLLAMA_BASE_URL: ${{ inputs.OLLAMA_BASE_URL }}
+    OLLAMA_MODEL: ${{ inputs.OLLAMA_MODEL }}
+    AMAZONQ_API_KEY: ${{ inputs.AMAZONQ_API_KEY }}
+    AMAZONQ_BASE_URL: ${{ inputs.AMAZONQ_BASE_URL }}
+    SONARQUBE_TOKEN: ${{ inputs.SONARQUBE_TOKEN }}
+    SONARQUBE_URL: ${{ inputs.SONARQUBE_URL }}
+    KILO_API_KEY: ${{ inputs.KILO_API_KEY }}
+    KILO_BASE_URL: ${{ inputs.KILO_BASE_URL }}
+    ROO_API_KEY: ${{ inputs.ROO_API_KEY }}
+    ROO_BASE_URL: ${{ inputs.ROO_BASE_URL }}
+    CLINE_API_KEY: ${{ inputs.CLINE_API_KEY }}
+    CLINE_BASE_URL: ${{ inputs.CLINE_BASE_URL }}
+    CONTINUE_API_KEY: ${{ inputs.CONTINUE_API_KEY }}
+    CONTINUE_BASE_URL: ${{ inputs.CONTINUE_BASE_URL }}
+    KADE_API_KEY: ${{ inputs.KADE_API_KEY }}
+    KADE_BASE_URL: ${{ inputs.KADE_BASE_URL }}
diff --git a/app/config.py b/app/config.py
new file mode 100644
index 0000000..d0f6c04
--- /dev/null
+++ b/app/config.py
@@ -0,0 +1,31 @@
+import os
+
+# Gitea
+GITEA_TOKEN = os.environ.get("GITEA_TOKEN", "")
+GITEA_SERVER_URL = os.environ.get("GITEA_SERVER_URL", "https://gitea.com")
+GITEA_REPOSITORY = os.environ.get("GITEA_REPOSITORY", "")
+PR_NUMBER = os.environ.get("PR_NUMBER", "")
+PR_HEAD_BRANCH = os.environ.get("PR_HEAD_BRANCH", "")
+PR_BASE_BRANCH = os.environ.get("PR_BASE_BRANCH", "")
+
+FINDINGS_PATH = ".gitea/ai-review/findings.json"
+
+
+def get_llm_config():
+    """依優先順序偵測可用的 LLM，回傳 (provider, api_key, base_url, model)"""
+    checks = [
+        ("openai",   os.environ.get("OPENAI_API_KEY"),   os.environ.get("OPENAI_BASE_URL", "https://api.openai.com/v1"),   os.environ.get("OPENAI_MODEL", "gpt-4o-mini")),
+        ("claude",   os.environ.get("CLAUDE_API_KEY"),   os.environ.get("CLAUDE_BASE_URL", "https://api.anthropic.com/v1"), os.environ.get("CLAUDE_MODEL", "claude-3-haiku-20240307")),
+        ("gemini",   os.environ.get("GEMINI_API_KEY"),   os.environ.get("GEMINI_BASE_URL", "https://generativelanguage.googleapis.com/v1beta"), os.environ.get("GEMINI_MODEL", "gemini-1.5-flash")),
+        ("ollama",   "ollama",                            os.environ.get("OLLAMA_BASE_URL", ""),  os.environ.get("OLLAMA_MODEL", "")),
+        ("amazonq",  os.environ.get("AMAZONQ_API_KEY"),  os.environ.get("AMAZONQ_BASE_URL", "https://q.api.aws"), os.environ.get("OPENAI_MODEL", "amazon-q")),
+        ("kilo",     os.environ.get("KILO_API_KEY"),     os.environ.get("KILO_BASE_URL", "https://api.kilocode.com/v1"), os.environ.get("OPENAI_MODEL", "kilo-default")),
+        ("roo",      os.environ.get("ROO_API_KEY"),      os.environ.get("ROO_BASE_URL", "https://api.roocode.com/v1"), os.environ.get("OPENAI_MODEL", "roo-default")),
+        ("cline",    os.environ.get("CLINE_API_KEY"),    os.environ.get("CLINE_BASE_URL", "https://api.cline.dev/v1"), os.environ.get("OPENAI_MODEL", "cline-default")),
+        ("continue", os.environ.get("CONTINUE_API_KEY"), os.environ.get("CONTINUE_BASE_URL", "https://api.continue.dev/v1"), os.environ.get("OPENAI_MODEL", "continue-default")),
+        ("kade",     os.environ.get("KADE_API_KEY"),     os.environ.get("KADE_BASE_URL", "https://api.kade.dev/v1"), os.environ.get("OPENAI_MODEL", "kade-default")),
+    ]
+    for provider, key, base_url, model in checks:
+        if key and base_url:
+            return provider, key, base_url, model
+    return None, None, None, None
diff --git a/app/gitea.py b/app/gitea.py
new file mode 100644
index 0000000..4a1cd4c
--- /dev/null
+++ b/app/gitea.py
@@ -0,0 +1,26 @@
+import requests
+from config import GITEA_TOKEN, GITEA_SERVER_URL, GITEA_REPOSITORY, PR_NUMBER
+
+
+def _headers():
+    return {"Authorization": f"token {GITEA_TOKEN}", "Content-Type": "application/json"}
+
+
+def _api(path: str) -> str:
+    return f"{GITEA_SERVER_URL.rstrip('/')}/api/v1{path}"
+
+
+def get_pr_diff() -> str:
+    """取得 PR 的 git diff 內容"""
+    url = _api(f"/repos/{GITEA_REPOSITORY}/pulls/{PR_NUMBER}.diff")
+    resp = requests.get(url, headers=_headers(), timeout=60)
+    resp.raise_for_status()
+    return resp.text
+
+
+def post_comment(body: str) -> dict:
+    """在 PR 發布 comment"""
+    url = _api(f"/repos/{GITEA_REPOSITORY}/issues/{PR_NUMBER}/comments")
+    resp = requests.post(url, headers=_headers(), json={"body": body}, timeout=30)
+    resp.raise_for_status()
+    return resp.json()
diff --git a/app/llm.py b/app/llm.py
new file mode 100644
index 0000000..b15a8ae
--- /dev/null
+++ b/app/llm.py
@@ -0,0 +1,50 @@
+import json
+import requests
+from config import get_llm_config
+
+
+def chat(system_prompt: str, user_content: str) -> str:
+    """呼叫 LLM，回傳回應文字。失敗時拋出例外。"""
+    provider, api_key, base_url, model = get_llm_config()
+    if not provider:
+        raise RuntimeError("未設定任何 LLM API Key")
+
+    print(f"  [LLM] provider={provider} model={model}")
+
+    # 所有服務統一用 OpenAI-compatible chat completions 介面
+    url = f"{base_url.rstrip('/')}/chat/completions"
+    headers = {
+        "Content-Type": "application/json",
+        "Authorization": f"Bearer {api_key}",
+    }
+    # Claude 額外 header
+    if provider == "claude":
+        headers["anthropic-version"] = "2023-06-01"
+
+    payload = {
+        "model": model,
+        "messages": [
+            {"role": "system", "content": system_prompt},
+            {"role": "user", "content": user_content},
+        ],
+        "temperature": 0.2,
+    }
+
+    resp = requests.post(url, headers=headers, json=payload, timeout=120)
+    resp.raise_for_status()
+    return resp.json()["choices"][0]["message"]["content"]
+
+
+def chat_json(system_prompt: str, user_content: str) -> list:
+    """呼叫 LLM 並解析 JSON 陣列回應。失敗時回傳空陣列。"""
+    try:
+        text = chat(system_prompt, user_content)
+        # 去除可能的 markdown code block
+        text = text.strip()
+        if text.startswith("```"):
+            text = text.split("\n", 1)[1]
+            text = text.rsplit("```", 1)[0]
+        return json.loads(text)
+    except Exception as e:
+        print(f"  [LLM] 解析失敗: {e}")
+        return []
diff --git a/app/main.py b/app/main.py
new file mode 100644
index 0000000..2c7c725
--- /dev/null
+++ b/app/main.py
@@ -0,0 +1,75 @@
+import sys
+import traceback
+from config import GITEA_REPOSITORY, PR_NUMBER, PR_HEAD_BRANCH, PR_BASE_BRANCH, get_llm_config
+from roles import load_roles, get_role_intro
+from gitea import get_pr_diff, post_comment
+
+
+def main():
+    print("=" * 60)
+    print("🚀 Step1: Pipeline 啟動")
+    print(f"  repo={GITEA_REPOSITORY}  PR=#{PR_NUMBER}")
+    print(f"  {PR_HEAD_BRANCH} -> {PR_BASE_BRANCH}")
+
+    # 偵測 LLM
+    provider, _, base_url, model = get_llm_config()
+    if not provider:
+        print("❌ 未設定任何 LLM API Key，請檢查 action inputs")
+        sys.exit(1)
+    print(f"  LLM: provider={provider}  model={model}  base_url={base_url}")
+
+    # 載入角色
+    roles = load_roles()
+    print(f"  已載入 {len(roles)} 個角色: {[r['name'] for r in roles]}")
+
+    # 取得 PR diff
+    print("\n📋 Step1: 取得 PR Diff")
+    try:
+        diff = get_pr_diff()
+        print(f"  diff 長度: {len(diff)} 字元")
+    except Exception as e:
+        print(f"  ❌ 取得 diff 失敗: {e}")
+        sys.exit(1)
+
+    if not diff.strip():
+        print("  ⚠️  diff 為空，無需審查")
+        sys.exit(0)
+
+    # 發布角色介紹 comment
+    print("\n💬 Step1: 發布角色介紹 Comment")
+    try:
+        intro = get_role_intro(roles)
+        intro += f"\n\n> 🔍 服務：{provider}  模型：{model}"
+        post_comment(intro)
+        print("  ✅ 角色介紹 comment 發布成功")
+    except Exception as e:
+        print(f"  ⚠️  comment 發布失敗（繼續執行）: {e}")
+
+    print("\n📊 Step2: Findings 產生（待實作）")
+    print("  [stub] 各角色分析 diff...")
+
+    print("\n🔀 Step3: Findings 合併與去重（待實作）")
+    print("  [stub] 合併新舊 findings...")
+
+    print("\n📝 Step4: Findings 寫入與 Comment 發布（待實作）")
+    print("  [stub] 寫入 findings.json，發布 comment...")
+
+    print("\n💾 Step5: 記憶區 Commit/Push（待實作）")
+    print("  [stub] commit & push findings.json...")
+
+    print("\n🚦 Step6: 嚴重問題檢查（待實作）")
+    print("  [stub] 檢查 critical findings...")
+
+    print("\n✅ Pipeline 完成")
+    print("=" * 60)
+
+
+if __name__ == "__main__":
+    try:
+        main()
+    except SystemExit:
+        raise
+    except Exception:
+        print("❌ Runner failed:")
+        traceback.print_exc()
+        sys.exit(1)
diff --git a/app/prompts/roles/maintainability.yaml b/app/prompts/roles/maintainability.yaml
new file mode 100644
index 0000000..23f22d3
--- /dev/null
+++ b/app/prompts/roles/maintainability.yaml
@@ -0,0 +1,23 @@
+name: "Leo"
+role: "可維護性審查員"
+personality: "有遠見、重視長期維護成本，常常思考「六個月後的自己能看懂嗎？」"
+focus: "程式碼複雜度、模組化、重複程式碼、文件完整性、錯誤處理、可測試性"
+system_prompt: |
+  你是 Leo，一位重視長期維護成本的審查員。你的工作是審查程式碼的可維護性，包含複雜度、模組化、重複程式碼、文件完整性、錯誤處理。
+  
+  請分析以下 Git Diff，找出所有可維護性相關問題。
+  
+  回傳 JSON 陣列，每個問題格式如下：
+  {
+    "level": "critical|warning|info",
+    "role": "Leo",
+    "location": "檔案路徑:行號 或 檔案路徑",
+    "suggestion": "繁體中文的具體修改建議"
+  }
+  
+  等級定義：
+  - critical：嚴重影響可維護性，會造成技術債（如超長函式、完全無文件的公開 API）
+  - warning：建議改善的可維護性問題
+  - info：可選的改善建議
+  
+  只回傳 JSON 陣列，不要有其他文字。如果沒有問題，回傳空陣列 []。
diff --git a/app/prompts/roles/performance.yaml b/app/prompts/roles/performance.yaml
new file mode 100644
index 0000000..51f6249
--- /dev/null
+++ b/app/prompts/roles/performance.yaml
@@ -0,0 +1,23 @@
+name: "Zara"
+role: "效能優化專家"
+personality: "追求極致效能，對任何不必要的資源消耗都感到不舒服，喜歡用數據說話"
+focus: "時間複雜度、空間複雜度、資料庫查詢效率、快取策略、不必要的重複運算"
+system_prompt: |
+  你是 Zara，一位追求極致效能的優化專家。你的工作是審查程式碼的效能問題，包含時間複雜度、空間複雜度、資料庫查詢效率、快取策略。
+  
+  請分析以下 Git Diff，找出所有效能相關問題。
+  
+  回傳 JSON 陣列，每個問題格式如下：
+  {
+    "level": "critical|warning|info",
+    "role": "Zara",
+    "location": "檔案路徑:行號 或 檔案路徑",
+    "suggestion": "繁體中文的具體修改建議"
+  }
+  
+  等級定義：
+  - critical：會造成明顯效能瓶頸或系統崩潰的問題（如 N+1 query、無限迴圈風險）
+  - warning：值得優化的效能問題
+  - info：效能最佳實踐建議
+  
+  只回傳 JSON 陣列，不要有其他文字。如果沒有問題，回傳空陣列 []。
diff --git a/app/prompts/roles/security.yaml b/app/prompts/roles/security.yaml
new file mode 100644
index 0000000..3bc5d31
--- /dev/null
+++ b/app/prompts/roles/security.yaml
@@ -0,0 +1,23 @@
+name: "Rex"
+role: "資安審查員"
+personality: "謹慎、多疑、對任何潛在風險都保持高度警覺，寧可誤報也不放過漏洞"
+focus: "安全漏洞、注入攻擊、敏感資料洩漏、認證授權問題、依賴套件風險"
+system_prompt: |
+  你是 Rex，一位謹慎的資安審查員。你的工作是審查程式碼中的安全漏洞、注入攻擊風險、敏感資料洩漏、認證授權問題。
+  
+  請分析以下 Git Diff，找出所有安全相關問題。
+  
+  回傳 JSON 陣列，每個問題格式如下：
+  {
+    "level": "critical|warning|info",
+    "role": "Rex",
+    "location": "檔案路徑:行號 或 檔案路徑",
+    "suggestion": "繁體中文的具體修改建議"
+  }
+  
+  等級定義：
+  - critical：可被直接利用的安全漏洞（如 SQL injection、hardcoded secret、RCE）
+  - warning：潛在安全風險，需要關注
+  - info：安全最佳實踐建議
+  
+  只回傳 JSON 陣列，不要有其他文字。如果沒有問題，回傳空陣列 []。
diff --git a/app/prompts/roles/style.yaml b/app/prompts/roles/style.yaml
new file mode 100644
index 0000000..75955a4
--- /dev/null
+++ b/app/prompts/roles/style.yaml
@@ -0,0 +1,23 @@
+name: "Aria"
+role: "程式碼風格審查員"
+personality: "嚴謹、注重細節、對程式碼整潔度有高度要求，說話直接但不失禮貌"
+focus: "程式碼風格、命名規範、格式一致性、可讀性"
+system_prompt: |
+  你是 Aria，一位嚴謹的程式碼風格審查員。你的工作是審查程式碼的風格、命名規範、格式一致性與可讀性。
+  
+  請分析以下 Git Diff，找出所有風格相關問題。
+  
+  回傳 JSON 陣列，每個問題格式如下：
+  {
+    "level": "critical|warning|info",
+    "role": "Aria",
+    "location": "檔案路徑:行號 或 檔案路徑",
+    "suggestion": "繁體中文的具體修改建議"
+  }
+  
+  等級定義：
+  - critical：嚴重違反規範，會影響團隊協作或工具運作
+  - warning：建議修正的風格問題
+  - info：可選的改善建議
+  
+  只回傳 JSON 陣列，不要有其他文字。如果沒有問題，回傳空陣列 []。
diff --git a/app/prompts/roles/testing.yaml b/app/prompts/roles/testing.yaml
new file mode 100644
index 0000000..e83cf05
--- /dev/null
+++ b/app/prompts/roles/testing.yaml
@@ -0,0 +1,23 @@
+name: "Maya"
+role: "測試品質審查員"
+personality: "對測試覆蓋率有執念，相信沒有測試的程式碼等於沒有完成，溫和但堅持"
+focus: "測試覆蓋率、測試品質、邊界條件、錯誤情境測試、測試可讀性"
+system_prompt: |
+  你是 Maya，一位對測試品質有高度要求的審查員。你的工作是審查程式碼的測試覆蓋率、測試品質、邊界條件處理。
+  
+  請分析以下 Git Diff，找出所有測試相關問題。
+  
+  回傳 JSON 陣列，每個問題格式如下：
+  {
+    "level": "critical|warning|info",
+    "role": "Maya",
+    "location": "檔案路徑:行號 或 檔案路徑",
+    "suggestion": "繁體中文的具體修改建議"
+  }
+  
+  等級定義：
+  - critical：完全缺少測試的核心功能，或測試邏輯有嚴重錯誤
+  - warning：測試覆蓋不足或測試品質有待改善
+  - info：測試最佳實踐建議
+  
+  只回傳 JSON 陣列，不要有其他文字。如果沒有問題，回傳空陣列 []。
diff --git a/app/requirements.txt b/app/requirements.txt
new file mode 100644
index 0000000..1ebdae8
--- /dev/null
+++ b/app/requirements.txt
@@ -0,0 +1,3 @@
+requests==2.31.0
+pyyaml==6.0.1
+openai==1.12.0
diff --git a/app/roles.py b/app/roles.py
new file mode 100644
index 0000000..07d123c
--- /dev/null
+++ b/app/roles.py
@@ -0,0 +1,22 @@
+import os
+import yaml
+
+ROLES_DIR = "/action/app/prompts/roles"
+
+
+def load_roles() -> list[dict]:
+    """載入所有角色定義"""
+    roles = []
+    for fname in sorted(os.listdir(ROLES_DIR)):
+        if fname.endswith(".yaml"):
+            with open(os.path.join(ROLES_DIR, fname), "r", encoding="utf-8") as f:
+                roles.append(yaml.safe_load(f))
+    return roles
+
+
+def get_role_intro(roles: list[dict]) -> str:
+    """產生角色介紹文字（用於 comment）"""
+    lines = ["## 🤖 AI Code Review 團隊", ""]
+    for r in roles:
+        lines.append(f"- **{r['name']}** ({r['role']}): {r['personality']}")
+    return "\n".join(lines)
diff --git a/entrypoint.sh b/entrypoint.sh
index 378b552..ec0b7b7 100644
--- a/entrypoint.sh
+++ b/entrypoint.sh
@@ -1,11 +1,8 @@
 #!/bin/bash
+set -e
 
-echo "Gitea Server Url: $GITEA_SERVER_URL"
+echo "🚀 AI Code Review Action 啟動"
+echo "Repository: $GITEA_REPOSITORY"
+echo "PR: #$PR_NUMBER  ($PR_HEAD_BRANCH -> $PR_BASE_BRANCH)"
 
-echo "Gitea Repository: $GITEA_REPOSITORY"
-
-echo "Gitea Runner Token: $RUNNER_TOKEN"
-
-echo "Input Text: $INPUT_TEXT"
-
-echo "text=$INPUT_TEXT" >> "$GITHUB_OUTPUT"
\ No newline at end of file
+exec python /action/app/main.py

From ec1f6c96e7650c68517088259f4599dfd20fa21a Mon Sep 17 00:00:00 2001
From: Jeffery <jiantw83@yahoo.com>
Date: Mon, 11 May 2026 07:24:47 +0000
Subject: [PATCH 2/7] =?UTF-8?q?feat:=20=E9=9A=8E=E6=AE=B5=E4=B8=80=20-=20?=
 =?UTF-8?q?=E6=94=B9=E7=94=A8=20Node.js=20=E5=AF=A6=E4=BD=9C=E5=9F=BA?=
 =?UTF-8?q?=E6=9C=AC=E6=B5=81=E7=A8=8B=E9=AA=A8=E6=9E=B6?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

- Dockerfile: 改用 node:20-slim
- entrypoint.sh: 執行 app/main.js
- app/package.json: axios + js-yaml + openai
- app/config.js: 環境變數與 LLM 自動偵測（10 種服務）
- app/llm.js: OpenAI-compatible 統一介面
- app/gitea.js: PR diff 取得與 comment 發布
- app/roles.js: 從 prompts/roles/*.yaml 載入角色
- app/main.js: pipeline 骨架，log 每個主要階段
---
 Dockerfile           |  4 +--
 app/config.js        | 27 ++++++++++++++++
 app/config.py        | 31 ------------------
 app/gitea.js         | 15 +++++++++
 app/gitea.py         | 26 ---------------
 app/llm.js           | 33 +++++++++++++++++++
 app/llm.py           | 50 -----------------------------
 app/main.js          | 71 +++++++++++++++++++++++++++++++++++++++++
 app/main.py          | 75 --------------------------------------------
 app/package.json     | 10 ++++++
 app/requirements.txt |  3 --
 app/roles.js         | 20 ++++++++++++
 app/roles.py         | 22 -------------
 entrypoint.sh        |  2 +-
 14 files changed, 179 insertions(+), 210 deletions(-)
 create mode 100644 app/config.js
 delete mode 100644 app/config.py
 create mode 100644 app/gitea.js
 delete mode 100644 app/gitea.py
 create mode 100644 app/llm.js
 delete mode 100644 app/llm.py
 create mode 100644 app/main.js
 delete mode 100644 app/main.py
 create mode 100644 app/package.json
 delete mode 100644 app/requirements.txt
 create mode 100644 app/roles.js
 delete mode 100644 app/roles.py

diff --git a/Dockerfile b/Dockerfile
index 1ef0bb9..484bf92 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -1,4 +1,4 @@
-FROM python:3.11-slim
+FROM node:20-slim
 
 RUN apt-get update && apt-get install -y --no-install-recommends \
     git \
@@ -9,7 +9,7 @@ WORKDIR /action
 COPY app/ /action/app/
 COPY entrypoint.sh /entrypoint.sh
 
-RUN pip install --no-cache-dir -r /action/app/requirements.txt && \
+RUN cd /action/app && npm install && \
     chmod +x /entrypoint.sh
 
 ENTRYPOINT ["/entrypoint.sh"]
diff --git a/app/config.js b/app/config.js
new file mode 100644
index 0000000..ea20e2c
--- /dev/null
+++ b/app/config.js
@@ -0,0 +1,27 @@
+export const GITEA_TOKEN = process.env.GITEA_TOKEN || '';
+export const GITEA_SERVER_URL = process.env.GITEA_SERVER_URL || 'https://gitea.com';
+export const GITEA_REPOSITORY = process.env.GITEA_REPOSITORY || '';
+export const PR_NUMBER = process.env.PR_NUMBER || '';
+export const PR_HEAD_BRANCH = process.env.PR_HEAD_BRANCH || '';
+export const PR_BASE_BRANCH = process.env.PR_BASE_BRANCH || '';
+
+export const FINDINGS_PATH = '.gitea/ai-review/findings.json';
+
+export function getLLMConfig() {
+  const checks = [
+    ['openai',   process.env.OPENAI_API_KEY,   process.env.OPENAI_BASE_URL || 'https://api.openai.com/v1',   process.env.OPENAI_MODEL || 'gpt-4o-mini'],
+    ['claude',   process.env.CLAUDE_API_KEY,   process.env.CLAUDE_BASE_URL || 'https://api.anthropic.com/v1', process.env.CLAUDE_MODEL || 'claude-3-haiku-20240307'],
+    ['gemini',   process.env.GEMINI_API_KEY,   process.env.GEMINI_BASE_URL || 'https://generativelanguage.googleapis.com/v1beta', process.env.GEMINI_MODEL || 'gemini-1.5-flash'],
+    ['ollama',   'ollama',                      process.env.OLLAMA_BASE_URL,  process.env.OLLAMA_MODEL],
+    ['amazonq',  process.env.AMAZONQ_API_KEY,  process.env.AMAZONQ_BASE_URL || 'https://q.api.aws', process.env.OPENAI_MODEL || 'amazon-q'],
+    ['kilo',     process.env.KILO_API_KEY,     process.env.KILO_BASE_URL || 'https://api.kilocode.com/v1', process.env.OPENAI_MODEL || 'kilo-default'],
+    ['roo',      process.env.ROO_API_KEY,      process.env.ROO_BASE_URL || 'https://api.roocode.com/v1', process.env.OPENAI_MODEL || 'roo-default'],
+    ['cline',    process.env.CLINE_API_KEY,    process.env.CLINE_BASE_URL || 'https://api.cline.dev/v1', process.env.OPENAI_MODEL || 'cline-default'],
+    ['continue', process.env.CONTINUE_API_KEY, process.env.CONTINUE_BASE_URL || 'https://api.continue.dev/v1', process.env.OPENAI_MODEL || 'continue-default'],
+    ['kade',     process.env.KADE_API_KEY,     process.env.KADE_BASE_URL || 'https://api.kade.dev/v1', process.env.OPENAI_MODEL || 'kade-default'],
+  ];
+  for (const [provider, key, baseURL, model] of checks) {
+    if (key && baseURL) return { provider, apiKey: key, baseURL, model };
+  }
+  return { provider: null, apiKey: null, baseURL: null, model: null };
+}
diff --git a/app/config.py b/app/config.py
deleted file mode 100644
index d0f6c04..0000000
--- a/app/config.py
+++ /dev/null
@@ -1,31 +0,0 @@
-import os
-
-# Gitea
-GITEA_TOKEN = os.environ.get("GITEA_TOKEN", "")
-GITEA_SERVER_URL = os.environ.get("GITEA_SERVER_URL", "https://gitea.com")
-GITEA_REPOSITORY = os.environ.get("GITEA_REPOSITORY", "")
-PR_NUMBER = os.environ.get("PR_NUMBER", "")
-PR_HEAD_BRANCH = os.environ.get("PR_HEAD_BRANCH", "")
-PR_BASE_BRANCH = os.environ.get("PR_BASE_BRANCH", "")
-
-FINDINGS_PATH = ".gitea/ai-review/findings.json"
-
-
-def get_llm_config():
-    """依優先順序偵測可用的 LLM，回傳 (provider, api_key, base_url, model)"""
-    checks = [
-        ("openai",   os.environ.get("OPENAI_API_KEY"),   os.environ.get("OPENAI_BASE_URL", "https://api.openai.com/v1"),   os.environ.get("OPENAI_MODEL", "gpt-4o-mini")),
-        ("claude",   os.environ.get("CLAUDE_API_KEY"),   os.environ.get("CLAUDE_BASE_URL", "https://api.anthropic.com/v1"), os.environ.get("CLAUDE_MODEL", "claude-3-haiku-20240307")),
-        ("gemini",   os.environ.get("GEMINI_API_KEY"),   os.environ.get("GEMINI_BASE_URL", "https://generativelanguage.googleapis.com/v1beta"), os.environ.get("GEMINI_MODEL", "gemini-1.5-flash")),
-        ("ollama",   "ollama",                            os.environ.get("OLLAMA_BASE_URL", ""),  os.environ.get("OLLAMA_MODEL", "")),
-        ("amazonq",  os.environ.get("AMAZONQ_API_KEY"),  os.environ.get("AMAZONQ_BASE_URL", "https://q.api.aws"), os.environ.get("OPENAI_MODEL", "amazon-q")),
-        ("kilo",     os.environ.get("KILO_API_KEY"),     os.environ.get("KILO_BASE_URL", "https://api.kilocode.com/v1"), os.environ.get("OPENAI_MODEL", "kilo-default")),
-        ("roo",      os.environ.get("ROO_API_KEY"),      os.environ.get("ROO_BASE_URL", "https://api.roocode.com/v1"), os.environ.get("OPENAI_MODEL", "roo-default")),
-        ("cline",    os.environ.get("CLINE_API_KEY"),    os.environ.get("CLINE_BASE_URL", "https://api.cline.dev/v1"), os.environ.get("OPENAI_MODEL", "cline-default")),
-        ("continue", os.environ.get("CONTINUE_API_KEY"), os.environ.get("CONTINUE_BASE_URL", "https://api.continue.dev/v1"), os.environ.get("OPENAI_MODEL", "continue-default")),
-        ("kade",     os.environ.get("KADE_API_KEY"),     os.environ.get("KADE_BASE_URL", "https://api.kade.dev/v1"), os.environ.get("OPENAI_MODEL", "kade-default")),
-    ]
-    for provider, key, base_url, model in checks:
-        if key and base_url:
-            return provider, key, base_url, model
-    return None, None, None, None
diff --git a/app/gitea.js b/app/gitea.js
new file mode 100644
index 0000000..787f801
--- /dev/null
+++ b/app/gitea.js
@@ -0,0 +1,15 @@
+import axios from 'axios';
+import { GITEA_TOKEN, GITEA_SERVER_URL, GITEA_REPOSITORY, PR_NUMBER } from './config.js';
+
+const headers = () => ({ Authorization: `token ${GITEA_TOKEN}`, 'Content-Type': 'application/json' });
+const api = (path) => `${GITEA_SERVER_URL.replace(/\/$/, '')}/api/v1${path}`;
+
+export async function getPRDiff() {
+  const resp = await axios.get(api(`/repos/${GITEA_REPOSITORY}/pulls/${PR_NUMBER}.diff`), { headers: headers(), timeout: 60000 });
+  return resp.data;
+}
+
+export async function postComment(body) {
+  const resp = await axios.post(api(`/repos/${GITEA_REPOSITORY}/issues/${PR_NUMBER}/comments`), { body }, { headers: headers(), timeout: 30000 });
+  return resp.data;
+}
diff --git a/app/gitea.py b/app/gitea.py
deleted file mode 100644
index 4a1cd4c..0000000
--- a/app/gitea.py
+++ /dev/null
@@ -1,26 +0,0 @@
-import requests
-from config import GITEA_TOKEN, GITEA_SERVER_URL, GITEA_REPOSITORY, PR_NUMBER
-
-
-def _headers():
-    return {"Authorization": f"token {GITEA_TOKEN}", "Content-Type": "application/json"}
-
-
-def _api(path: str) -> str:
-    return f"{GITEA_SERVER_URL.rstrip('/')}/api/v1{path}"
-
-
-def get_pr_diff() -> str:
-    """取得 PR 的 git diff 內容"""
-    url = _api(f"/repos/{GITEA_REPOSITORY}/pulls/{PR_NUMBER}.diff")
-    resp = requests.get(url, headers=_headers(), timeout=60)
-    resp.raise_for_status()
-    return resp.text
-
-
-def post_comment(body: str) -> dict:
-    """在 PR 發布 comment"""
-    url = _api(f"/repos/{GITEA_REPOSITORY}/issues/{PR_NUMBER}/comments")
-    resp = requests.post(url, headers=_headers(), json={"body": body}, timeout=30)
-    resp.raise_for_status()
-    return resp.json()
diff --git a/app/llm.js b/app/llm.js
new file mode 100644
index 0000000..f2521d3
--- /dev/null
+++ b/app/llm.js
@@ -0,0 +1,33 @@
+import axios from 'axios';
+import { getLLMConfig } from './config.js';
+
+export async function chat(systemPrompt, userContent) {
+  const { provider, apiKey, baseURL, model } = getLLMConfig();
+  if (!provider) throw new Error('未設定任何 LLM API Key');
+
+  console.log(`  [LLM] provider=${provider} model=${model}`);
+
+  const headers = {
+    'Content-Type': 'application/json',
+    'Authorization': `Bearer ${apiKey}`,
+  };
+  if (provider === 'claude') headers['anthropic-version'] = '2023-06-01';
+
+  const resp = await axios.post(
+    `${baseURL.replace(/\/$/, '')}/chat/completions`,
+    { model, messages: [{ role: 'system', content: systemPrompt }, { role: 'user', content: userContent }], temperature: 0.2 },
+    { headers, timeout: 120000 }
+  );
+  return resp.data.choices[0].message.content;
+}
+
+export async function chatJSON(systemPrompt, userContent) {
+  try {
+    let text = await chat(systemPrompt, userContent);
+    text = text.trim().replace(/^```[^\n]*\n?/, '').replace(/```$/, '').trim();
+    return JSON.parse(text);
+  } catch (e) {
+    console.log(`  [LLM] 解析失敗: ${e.message}`);
+    return [];
+  }
+}
diff --git a/app/llm.py b/app/llm.py
deleted file mode 100644
index b15a8ae..0000000
--- a/app/llm.py
+++ /dev/null
@@ -1,50 +0,0 @@
-import json
-import requests
-from config import get_llm_config
-
-
-def chat(system_prompt: str, user_content: str) -> str:
-    """呼叫 LLM，回傳回應文字。失敗時拋出例外。"""
-    provider, api_key, base_url, model = get_llm_config()
-    if not provider:
-        raise RuntimeError("未設定任何 LLM API Key")
-
-    print(f"  [LLM] provider={provider} model={model}")
-
-    # 所有服務統一用 OpenAI-compatible chat completions 介面
-    url = f"{base_url.rstrip('/')}/chat/completions"
-    headers = {
-        "Content-Type": "application/json",
-        "Authorization": f"Bearer {api_key}",
-    }
-    # Claude 額外 header
-    if provider == "claude":
-        headers["anthropic-version"] = "2023-06-01"
-
-    payload = {
-        "model": model,
-        "messages": [
-            {"role": "system", "content": system_prompt},
-            {"role": "user", "content": user_content},
-        ],
-        "temperature": 0.2,
-    }
-
-    resp = requests.post(url, headers=headers, json=payload, timeout=120)
-    resp.raise_for_status()
-    return resp.json()["choices"][0]["message"]["content"]
-
-
-def chat_json(system_prompt: str, user_content: str) -> list:
-    """呼叫 LLM 並解析 JSON 陣列回應。失敗時回傳空陣列。"""
-    try:
-        text = chat(system_prompt, user_content)
-        # 去除可能的 markdown code block
-        text = text.strip()
-        if text.startswith("```"):
-            text = text.split("\n", 1)[1]
-            text = text.rsplit("```", 1)[0]
-        return json.loads(text)
-    except Exception as e:
-        print(f"  [LLM] 解析失敗: {e}")
-        return []
diff --git a/app/main.js b/app/main.js
new file mode 100644
index 0000000..16981ba
--- /dev/null
+++ b/app/main.js
@@ -0,0 +1,71 @@
+import { GITEA_REPOSITORY, PR_NUMBER, PR_HEAD_BRANCH, PR_BASE_BRANCH, getLLMConfig } from './config.js';
+import { loadRoles, getRoleIntro } from './roles.js';
+import { getPRDiff, postComment } from './gitea.js';
+
+async function main() {
+  console.log('='.repeat(60));
+  console.log('🚀 Step1: Pipeline 啟動');
+  console.log(`  repo=${GITEA_REPOSITORY}  PR=#${PR_NUMBER}`);
+  console.log(`  ${PR_HEAD_BRANCH} -> ${PR_BASE_BRANCH}`);
+
+  // 偵測 LLM
+  const { provider, baseURL, model } = getLLMConfig();
+  if (!provider) {
+    console.error('❌ 未設定任何 LLM API Key，請檢查 action inputs');
+    process.exit(1);
+  }
+  console.log(`  LLM: provider=${provider}  model=${model}  base_url=${baseURL}`);
+
+  // 載入角色
+  const roles = loadRoles();
+  console.log(`  已載入 ${roles.length} 個角色: [${roles.map(r => r.name).join(', ')}]`);
+
+  // 取得 PR diff
+  console.log('\n📋 Step1: 取得 PR Diff');
+  let diff;
+  try {
+    diff = await getPRDiff();
+    console.log(`  diff 長度: ${diff.length} 字元`);
+  } catch (e) {
+    console.error(`  ❌ 取得 diff 失敗: ${e.message}`);
+    process.exit(1);
+  }
+
+  if (!diff.trim()) {
+    console.log('  ⚠️  diff 為空，無需審查');
+    process.exit(0);
+  }
+
+  // 發布角色介紹 comment
+  console.log('\n💬 Step1: 發布角色介紹 Comment');
+  try {
+    const intro = getRoleIntro(roles) + `\n\n> 🔍 服務：${provider}  模型：${model}`;
+    await postComment(intro);
+    console.log('  ✅ 角色介紹 comment 發布成功');
+  } catch (e) {
+    console.log(`  ⚠️  comment 發布失敗（繼續執行）: ${e.message}`);
+  }
+
+  console.log('\n📊 Step2: Findings 產生（待實作）');
+  console.log('  [stub] 各角色分析 diff...');
+
+  console.log('\n🔀 Step3: Findings 合併與去重（待實作）');
+  console.log('  [stub] 合併新舊 findings...');
+
+  console.log('\n📝 Step4: Findings 寫入與 Comment 發布（待實作）');
+  console.log('  [stub] 寫入 findings.json，發布 comment...');
+
+  console.log('\n💾 Step5: 記憶區 Commit/Push（待實作）');
+  console.log('  [stub] commit & push findings.json...');
+
+  console.log('\n🚦 Step6: 嚴重問題檢查（待實作）');
+  console.log('  [stub] 檢查 critical findings...');
+
+  console.log('\n✅ Pipeline 完成');
+  console.log('='.repeat(60));
+}
+
+main().catch(e => {
+  console.error('❌ Runner failed:', e.message);
+  process.exit(1);
+});
diff --git a/app/main.py b/app/main.py
deleted file mode 100644
index 2c7c725..0000000
--- a/app/main.py
+++ /dev/null
@@ -1,75 +0,0 @@
-import sys
-import traceback
-from config import GITEA_REPOSITORY, PR_NUMBER, PR_HEAD_BRANCH, PR_BASE_BRANCH, get_llm_config
-from roles import load_roles, get_role_intro
-from gitea import get_pr_diff, post_comment
-
-
-def main():
-    print("=" * 60)
-    print("🚀 Step1: Pipeline 啟動")
-    print(f"  repo={GITEA_REPOSITORY}  PR=#{PR_NUMBER}")
-    print(f"  {PR_HEAD_BRANCH} -> {PR_BASE_BRANCH}")
-
-    # 偵測 LLM
-    provider, _, base_url, model = get_llm_config()
-    if not provider:
-        print("❌ 未設定任何 LLM API Key，請檢查 action inputs")
-        sys.exit(1)
-    print(f"  LLM: provider={provider}  model={model}  base_url={base_url}")
-
-    # 載入角色
-    roles = load_roles()
-    print(f"  已載入 {len(roles)} 個角色: {[r['name'] for r in roles]}")
-
-    # 取得 PR diff
-    print("\n📋 Step1: 取得 PR Diff")
-    try:
-        diff = get_pr_diff()
-        print(f"  diff 長度: {len(diff)} 字元")
-    except Exception as e:
-        print(f"  ❌ 取得 diff 失敗: {e}")
-        sys.exit(1)
-
-    if not diff.strip():
-        print("  ⚠️  diff 為空，無需審查")
-        sys.exit(0)
-
-    # 發布角色介紹 comment
-    print("\n💬 Step1: 發布角色介紹 Comment")
-    try:
-        intro = get_role_intro(roles)
-        intro += f"\n\n> 🔍 服務：{provider}  模型：{model}"
-        post_comment(intro)
-        print("  ✅ 角色介紹 comment 發布成功")
-    except Exception as e:
-        print(f"  ⚠️  comment 發布失敗（繼續執行）: {e}")
-
-    print("\n📊 Step2: Findings 產生（待實作）")
-    print("  [stub] 各角色分析 diff...")
-
-    print("\n🔀 Step3: Findings 合併與去重（待實作）")
-    print("  [stub] 合併新舊 findings...")
-
-    print("\n📝 Step4: Findings 寫入與 Comment 發布（待實作）")
-    print("  [stub] 寫入 findings.json，發布 comment...")
-
-    print("\n💾 Step5: 記憶區 Commit/Push（待實作）")
-    print("  [stub] commit & push findings.json...")
-
-    print("\n🚦 Step6: 嚴重問題檢查（待實作）")
-    print("  [stub] 檢查 critical findings...")
-
-    print("\n✅ Pipeline 完成")
-    print("=" * 60)
-
-
-if __name__ == "__main__":
-    try:
-        main()
-    except SystemExit:
-        raise
-    except Exception:
-        print("❌ Runner failed:")
-        traceback.print_exc()
-        sys.exit(1)
diff --git a/app/package.json b/app/package.json
new file mode 100644
index 0000000..b5e3877
--- /dev/null
+++ b/app/package.json
@@ -0,0 +1,10 @@
+{
+  "name": "ai-code-review",
+  "version": "1.0.0",
+  "type": "module",
+  "dependencies": {
+    "axios": "^1.6.7",
+    "js-yaml": "^4.1.0",
+    "openai": "^4.28.0"
+  }
+}
diff --git a/app/requirements.txt b/app/requirements.txt
deleted file mode 100644
index 1ebdae8..0000000
--- a/app/requirements.txt
+++ /dev/null
@@ -1,3 +0,0 @@
-requests==2.31.0
-pyyaml==6.0.1
-openai==1.12.0
diff --git a/app/roles.js b/app/roles.js
new file mode 100644
index 0000000..f36e2d4
--- /dev/null
+++ b/app/roles.js
@@ -0,0 +1,20 @@
+import fs from 'fs';
+import path from 'path';
+import yaml from 'js-yaml';
+
+const ROLES_DIR = '/action/app/prompts/roles';
+
+export function loadRoles() {
+  return fs.readdirSync(ROLES_DIR)
+    .filter(f => f.endsWith('.yaml'))
+    .sort()
+    .map(f => yaml.load(fs.readFileSync(path.join(ROLES_DIR, f), 'utf8')));
+}
+
+export function getRoleIntro(roles) {
+  const lines = ['## 🤖 AI Code Review 團隊', ''];
+  for (const r of roles) {
+    lines.push(`- **${r.name}** (${r.role})：${r.personality}`);
+  }
+  return lines.join('\n');
+}
diff --git a/app/roles.py b/app/roles.py
deleted file mode 100644
index 07d123c..0000000
--- a/app/roles.py
+++ /dev/null
@@ -1,22 +0,0 @@
-import os
-import yaml
-
-ROLES_DIR = "/action/app/prompts/roles"
-
-
-def load_roles() -> list[dict]:
-    """載入所有角色定義"""
-    roles = []
-    for fname in sorted(os.listdir(ROLES_DIR)):
-        if fname.endswith(".yaml"):
-            with open(os.path.join(ROLES_DIR, fname), "r", encoding="utf-8") as f:
-                roles.append(yaml.safe_load(f))
-    return roles
-
-
-def get_role_intro(roles: list[dict]) -> str:
-    """產生角色介紹文字（用於 comment）"""
-    lines = ["## 🤖 AI Code Review 團隊", ""]
-    for r in roles:
-        lines.append(f"- **{r['name']}** ({r['role']}): {r['personality']}")
-    return "\n".join(lines)
diff --git a/entrypoint.sh b/entrypoint.sh
index ec0b7b7..64c4ccb 100644
--- a/entrypoint.sh
+++ b/entrypoint.sh
@@ -5,4 +5,4 @@ echo "🚀 AI Code Review Action 啟動"
 echo "Repository: $GITEA_REPOSITORY"
 echo "PR: #$PR_NUMBER  ($PR_HEAD_BRANCH -> $PR_BASE_BRANCH)"
 
-exec python /action/app/main.py
+exec node /action/app/main.js

From bccf0e5b0b7e74b134f3d0f656c31601e81b1452 Mon Sep 17 00:00:00 2001
From: Jeffery <jiantw83@yahoo.com>
Date: Mon, 11 May 2026 07:29:52 +0000
Subject: [PATCH 3/7] feat: add AI code review workflow configuration

Co-authored-by: Copilot <copilot@github.com>
---
 .gitea/workflows/review.yaml | 18 ++++++++++++++++++
 1 file changed, 18 insertions(+)
 create mode 100644 .gitea/workflows/review.yaml

diff --git a/.gitea/workflows/review.yaml b/.gitea/workflows/review.yaml
new file mode 100644
index 0000000..6612336
--- /dev/null
+++ b/.gitea/workflows/review.yaml
@@ -0,0 +1,18 @@
+name: AI
+on:
+  pull_request:
+    types: [opened, synchronize]
+jobs:
+  code-review:
+    name: 'Code Review'
+    runs-on: ubuntu
+    steps:
+    - name: AI Code Review
+      uses: ../../action.yaml
+      with:
+        OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
+        OPENAI_BASE_URL: https://api.openai.com/v1
+    permissions:
+      contents: write
+      pull-requests: write
+      issues: write
\ No newline at end of file

From 43e21d07cdb231f1116cbd18870b152262fdef2f Mon Sep 17 00:00:00 2001
From: Jeffery <jiantw83@yahoo.com>
Date: Mon, 11 May 2026 07:37:04 +0000
Subject: [PATCH 4/7] feat: update AI code review workflow to use specific
 action URL and add tagging step

Co-authored-by: Copilot <copilot@github.com>
---
 .gitea/workflows/review.yaml | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/.gitea/workflows/review.yaml b/.gitea/workflows/review.yaml
index 6612336..0d8f2c1 100644
--- a/.gitea/workflows/review.yaml
+++ b/.gitea/workflows/review.yaml
@@ -7,8 +7,15 @@ jobs:
     name: 'Code Review'
     runs-on: ubuntu
     steps:
+    - name: Tag develop on source branch
+      uses: akkuman/gitea-release-action@${{ vars.ACTION_RELEASE_VERSION }}
+      with:
+        tag: develop
+        target: ${{ github.head_ref }}
+        generate_release_notes: false
+        prerelease: false
     - name: AI Code Review
-      uses: ../../action.yaml
+      uses: https://gitea.jsc.idv.tw/jiantw83/code-review@develop
       with:
         OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
         OPENAI_BASE_URL: https://api.openai.com/v1

From fa5a734166a33238519676eddf289d5e7e76e48f Mon Sep 17 00:00:00 2001
From: Jeffery <jiantw83@yahoo.com>
Date: Mon, 11 May 2026 07:39:29 +0000
Subject: [PATCH 5/7] fix: ensure 'name' parameter is included in the tag step
 of the code review workflow

Co-authored-by: Copilot <copilot@github.com>
---
 .gitea/workflows/review.yaml | 1 +
 1 file changed, 1 insertion(+)

diff --git a/.gitea/workflows/review.yaml b/.gitea/workflows/review.yaml
index 0d8f2c1..e07f83c 100644
--- a/.gitea/workflows/review.yaml
+++ b/.gitea/workflows/review.yaml
@@ -10,6 +10,7 @@ jobs:
     - name: Tag develop on source branch
       uses: akkuman/gitea-release-action@${{ vars.ACTION_RELEASE_VERSION }}
       with:
+        name: develop
         tag: develop
         target: ${{ github.head_ref }}
         generate_release_notes: false

From cdac64e22400f7c783bf218de859ad0ef164063a Mon Sep 17 00:00:00 2001
From: Jeffery <jiantw83@yahoo.com>
Date: Mon, 11 May 2026 07:46:59 +0000
Subject: [PATCH 6/7] feat: refactor code review workflow to remove tagging
 step and use dynamic branch reference

---
 .gitea/workflows/review.yaml | 10 +---------
 1 file changed, 1 insertion(+), 9 deletions(-)

diff --git a/.gitea/workflows/review.yaml b/.gitea/workflows/review.yaml
index e07f83c..18e356e 100644
--- a/.gitea/workflows/review.yaml
+++ b/.gitea/workflows/review.yaml
@@ -7,16 +7,8 @@ jobs:
     name: 'Code Review'
     runs-on: ubuntu
     steps:
-    - name: Tag develop on source branch
-      uses: akkuman/gitea-release-action@${{ vars.ACTION_RELEASE_VERSION }}
-      with:
-        name: develop
-        tag: develop
-        target: ${{ github.head_ref }}
-        generate_release_notes: false
-        prerelease: false
     - name: AI Code Review
-      uses: https://gitea.jsc.idv.tw/jiantw83/code-review@develop
+      uses: https://gitea.jsc.idv.tw/jiantw83/code-review@${{ github.head_ref }}
       with:
         OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
         OPENAI_BASE_URL: https://api.openai.com/v1

From 46dd8320d143524309d8bdc284463bd5739cc991 Mon Sep 17 00:00:00 2001
From: Jeffery <jiantw83@yahoo.com>
Date: Mon, 11 May 2026 07:52:21 +0000
Subject: [PATCH 7/7] =?UTF-8?q?feat:=20=E9=9A=8E=E6=AE=B5=E4=BA=8C=20-=20F?=
 =?UTF-8?q?indings=20=E7=94=A2=E7=94=9F=E8=88=87=E5=90=88=E4=BD=B5?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

- app/findings.js: 各角色分析 diff、讀取舊 findings、合併去重、等級排序
- app/main.js: 實作 Step2/Step3，log findings 統計
---
 app/findings.js | 63 +++++++++++++++++++++++++++++++++++++++++++++++++
 app/main.js     | 25 ++++++++++++++++----
 2 files changed, 84 insertions(+), 4 deletions(-)
 create mode 100644 app/findings.js

diff --git a/app/findings.js b/app/findings.js
new file mode 100644
index 0000000..3975a80
--- /dev/null
+++ b/app/findings.js
@@ -0,0 +1,63 @@
+import fs from 'fs';
+import path from 'path';
+import { chatJSON } from './llm.js';
+import { FINDINGS_PATH } from './config.js';
+
+const LEVELS = ['critical', 'warning', 'info'];
+
+/**
+ * 用單一角色分析 diff，回傳 findings 陣列
+ */
+export async function analyzeWithRole(role, diff) {
+  console.log(`  [${role.name}] 開始分析...`);
+  const findings = await chatJSON(role.system_prompt, `以下是 Git Diff 內容：\n\n${diff}`);
+  // 確保每筆都有必要欄位，並標記為新問題
+  const valid = findings.filter(f => f.level && f.role && f.location && f.suggestion)
+    .map(f => ({ ...f, is_new: true }));
+  console.log(`  [${role.name}] 找到 ${valid.length} 個問題`);
+  return valid;
+}
+
+/**
+ * 讀取舊 findings（從 workspace 的 FINDINGS_PATH）
+ */
+export function loadOldFindings(workspace) {
+  const fullPath = path.join(workspace, FINDINGS_PATH);
+  if (!fs.existsSync(fullPath)) {
+    console.log('  舊 findings 檔案不存在，視為空');
+    return [];
+  }
+  try {
+    const data = JSON.parse(fs.readFileSync(fullPath, 'utf8'));
+    const old = (Array.isArray(data) ? data : []).map(f => ({ ...f, is_new: false }));
+    console.log(`  讀取舊 findings: ${old.length} 筆`);
+    return old;
+  } catch (e) {
+    console.log(`  ⚠️  讀取舊 findings 失敗: ${e.message}，視為空`);
+    return [];
+  }
+}
+
+/**
+ * 合併新舊 findings，以 (role + location + suggestion前50字) 為 key 去除重複
+ * 舊問題保留，新問題若與舊問題重複則捨棄
+ */
+export function mergeFindings(oldFindings, newFindings) {
+  const key = f => `${f.role}|${f.location}|${String(f.suggestion).slice(0, 50)}`;
+  const seen = new Set(oldFindings.map(key));
+  const deduped = newFindings.filter(f => {
+    if (seen.has(key(f))) return false;
+    seen.add(key(f));
+    return true;
+  });
+  const merged = [...oldFindings, ...deduped];
+  console.log(`  合併結果: 舊=${oldFindings.length} 新(去重後)=${deduped.length} 總計=${merged.length}`);
+  return merged;
+}
+
+/**
+ * 依等級排序（critical > warning > info）
+ */
+export function sortByLevel(findings) {
+  return [...findings].sort((a, b) => LEVELS.indexOf(a.level) - LEVELS.indexOf(b.level));
+}
diff --git a/app/main.js b/app/main.js
index 16981ba..e317d61 100644
--- a/app/main.js
+++ b/app/main.js
@@ -1,6 +1,9 @@
 import { GITEA_REPOSITORY, PR_NUMBER, PR_HEAD_BRANCH, PR_BASE_BRANCH, getLLMConfig } from './config.js';
 import { loadRoles, getRoleIntro } from './roles.js';
 import { getPRDiff, postComment } from './gitea.js';
+import { analyzeWithRole, loadOldFindings, mergeFindings, sortByLevel } from './findings.js';
+
+const WORKSPACE = process.env.GITHUB_WORKSPACE || '/workspace';
 
 async function main() {
   console.log('='.repeat(60));
@@ -46,11 +49,25 @@ async function main() {
     console.log(`  ⚠️  comment 發布失敗（繼續執行）: ${e.message}`);
   }
 
-  console.log('\n📊 Step2: Findings 產生（待實作）');
-  console.log('  [stub] 各角色分析 diff...');
+  // Step2: 各角色分析 diff 產生新 findings
+  console.log('\n📊 Step2: Findings 產生');
+  const newFindings = [];
+  for (const role of roles) {
+    try {
+      const found = await analyzeWithRole(role, diff);
+      newFindings.push(...found);
+    } catch (e) {
+      console.log(`  ⚠️  [${role.name}] 分析失敗（跳過）: ${e.message}`);
+    }
+  }
+  console.log(`  Step2 完成: 新 findings 總計 ${newFindings.length} 筆`);
 
-  console.log('\n🔀 Step3: Findings 合併與去重（待實作）');
-  console.log('  [stub] 合併新舊 findings...');
+  // Step3: 讀取舊 findings，合併去重
+  console.log('\n🔀 Step3: Findings 合併');
+  const oldFindings = loadOldFindings(WORKSPACE);
+  const mergedFindings = mergeFindings(oldFindings, newFindings);
+  const sorted = sortByLevel(mergedFindings);
+  console.log(`  Step3 merged findings total=${sorted.length} (critical=${sorted.filter(f=>f.level==='critical').length} warning=${sorted.filter(f=>f.level==='warning').length} info=${sorted.filter(f=>f.level==='info').length})`);
 
   console.log('\n📝 Step4: Findings 寫入與 Comment 發布（待實作）');
   console.log('  [stub] 寫入 findings.json，發布 comment...');