一键导入
dual-code-review
Orchestrate a comprehensive code review process involving both Gemini (Antigravity) and Claude CLI, synthesizing a consensus report.
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。
菜单
Orchestrate a comprehensive code review process involving both Gemini (Antigravity) and Claude CLI, synthesizing a consensus report.
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
Apple Search Ads (ASA) deep analysis for mobile app advertisers. Evaluates campaign structure, bid health, Creative Sets, MMP attribution, budget pacing, TAP coverage (Today/Search/Product Pages), and goal CPA benchmarks by country. Use when user says Apple Search Ads, ASA, App Store ads, Apple ads, Search Ads, or is advertising a mobile app on iOS.
Competitor ad intelligence analysis across Google, Meta, LinkedIn, TikTok, and Microsoft. Analyzes competitor ad copy, creative strategy, keyword targeting, estimated spend, and identifies competitive gaps and opportunities. Use when user says competitor ads, ad spy, competitive analysis, competitor PPC, or ad intelligence.
Campaign concept and copy brief generator for paid advertising. Reads brand-profile.json and optional audit results to produce structured campaign concepts, messaging pillars, and copy briefs. Outputs campaign-brief.md to the current directory. Run after /ads dna and before /ads generate. Triggers on: create campaign, campaign brief, ad concepts, write ad copy, campaign strategy, ad messaging, creative brief, generate concepts.
Brand DNA extractor for paid advertising. Scans a website URL to extract visual identity, tone of voice, color palette, typography, and imagery style. Outputs brand-profile.json to the current directory. Run before /ads create or /ads generate for brand-consistent creative. Triggers on: brand DNA, brand profile, extract brand, brand identity, brand colors, what is the brand voice, analyze brand, brand style guide.
AI image generation for paid ad creatives. Reads campaign-brief.md and brand-profile.json to produce platform-sized ad images using banana-claude. Requires banana-claude (v1.4.1+) with nanobanana-mcp configured. Triggers on: generate ads, create images, make ad creatives, generate visuals, create ad images, generate campaign images, make the images, generate from brief.
Landing page quality assessment for paid advertising campaigns. Evaluates message match, page speed, mobile experience, trust signals, form optimization, and conversion rate potential. Use when user says landing page, post-click experience, landing page audit, conversion rate, or landing page optimization.
| name | dual-code-review |
| description | Orchestrate a comprehensive code review process involving both Gemini (Antigravity) and Claude CLI, synthesizing a consensus report. |
透過 Gemini 與 Claude CLI 的雙重審查,產出高可信度的程式碼審查報告。
flowchart TD
START([開始審查]) --> SCOPE[確定審查範圍]
SCOPE --> PARALLEL{平行分析}
PARALLEL --> GEMINI[Antigravity 分析]
PARALLEL --> CLAUDE[Claude CLI 分析]
GEMINI --> COLLECT[收集兩方結果]
CLAUDE --> COLLECT
COLLECT --> COMPARE{比對發現}
COMPARE -->|雙方都發現| CONFIRMED[✅ Double Verified]
COMPARE -->|僅 Claude 發現| VERIFY_C{驗證 Claude 發現}
COMPARE -->|僅 Gemini 發現| VERIFY_G{驗證 Gemini 發現}
VERIFY_C -->|有效| ACCEPT_C[接受並標記]
VERIFY_C -->|幻覺| DISCARD_C[捨棄]
VERIFY_G -->|有效| ACCEPT_G[接受並標記]
CONFIRMED --> REPORT[產出報告]
ACCEPT_C --> REPORT
ACCEPT_G --> REPORT
DISCARD_C --> REPORT
REPORT --> DONE([完成])
# 選項 A: 暫存區變更
git diff --staged > /tmp/review_content.txt
# 選項 B: 與 main 分支比較
git diff main > /tmp/review_content.txt
# 選項 C: 特定檔案
cat src/auth/*.ts > /tmp/review_content.txt
# 唯讀審查,強制 JSON 輸出
cat /tmp/review_content.txt | claude -p \
"你是資深程式碼審查員。請審查以下程式碼變更。
重點關注:Bug、安全性、效能、可讀性。
以 JSON 格式回傳發現。" \
--output-format json \
--json-schema '{
"type": "object",
"properties": {
"findings": {
"type": "array",
"items": {
"type": "object",
"properties": {
"severity": {"type": "string", "enum": ["High", "Medium", "Low"]},
"category": {"type": "string"},
"file": {"type": "string"},
"line": {"type": "integer"},
"message": {"type": "string"},
"suggestion": {"type": "string"}
},
"required": ["severity", "category", "message"]
}
},
"summary": {"type": "string"}
}
}' \
--allowedTools "Read,Grep,Glob" \
--max-turns 5 | jq '.structured_output' > /tmp/claude_review.json
Gemini 同步進行獨立分析,在讀取 Claude 結果前先完成自己的判斷,以維持獨立性。
評估邏輯:
| 情況 | 處理方式 | 標籤 |
|---|---|---|
| 雙方都發現相同問題 | 確認為高可信度問題 | [Double Verified] |
| 僅 Claude 發現 | 驗證是否為幻覺,有效則接受 | [Claude CLI Detected] |
| 僅 Gemini 發現 | 再次確認,有效則保留 | [Antigravity Detected] |
報告結構:
# Code Review Report
## Summary
- **Status**: Pass / Request Changes
- **Files Reviewed**: X
- **Issues Found**: High: X, Medium: X, Low: X
## Consensus Highlights
> 雙方一致認為的關鍵問題
## Detailed Findings
### 🔴 High Severity
| Issue | File | Line | Source |
|-------|------|------|--------|
| ... | ... | ... | [Double Verified] |
### 🟡 Medium Severity
...
### 🟢 Low Severity
...
## Actionable Advice
1. 修復步驟...
# 階段一:唯讀審查
review=$(git diff --staged | claude -p \
"審查這些變更,列出所有問題" \
--allowedTools "Read,Grep,Glob" \
--output-format json)
# 階段二:自動修復(需人工確認)
echo "$review" | jq -r '.result' | claude -p \
"依照以下審查結果修復問題:$(cat)" \
--allowedTools "Read,Edit,Write" \
--max-turns 10
當使用者提及以下關鍵字時啟用此技能: