Back to Details

nutmeg-review

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Review

审查

Dispatch specialised reviewers to check football data code and visualisations for correctness, convention compliance, and edge cases.
调度专门的审查员来校验足球数据代码及可视化内容的正确性、规范合规性以及边界场景问题。

Accuracy

准确性

Read and follow 
docs/accuracy-guardrail.md
before answering any question about provider-specific facts.
在回答任何与数据提供商相关的事实问题前，请阅读并遵循
docs/accuracy-guardrail.md
的要求。

First: check profile

第一步：检查用户配置

Read 
.nutmeg.user.md
. If it doesn't exist, tell the user to run 
/nutmeg
first.
读取
.nutmeg.user.md
文件。如果该文件不存在，告知用户需要先运行
/nutmeg
命令。

Determine scope

确定审查范围

Look at what the user wants reviewed. Read the relevant files. Then decide which reviewers to dispatch:
SignalDispatch
Code processes football data (fetching, filtering, transforming, computing metrics)data-reviewer agent
Code renders a chart or visualisationchart-reviewer agent (Mode 1: Code Review)
User provides a URL or says "check how it looks"chart-reviewer agent (Mode 2: Visual Inspection)
Chart has filters, tooltips, state, or dynamic datachart-reviewer agent (Mode 3: Interactive Edge Cases)
Code does both data processing AND chart renderingBoth agents in parallel
Always dispatch at least one. If unclear, dispatch both — redundant findings are better than missed issues.
查看用户需要审查的内容，读取相关文件，然后决定需要调度哪些审查员：
触发信号调度的审查员
代码处理足球数据（抓取、过滤、转换、指标计算）data-reviewer agent
代码渲染图表或可视化内容chart-reviewer agent（模式1：代码审查）
用户提供了URL或者表示“检查下展示效果”chart-reviewer agent（模式2：视觉检查）
图表包含过滤器、提示框、状态或者动态数据chart-reviewer agent（模式3：交互边界场景测试）
代码同时包含数据处理和图表渲染逻辑并行调用两个agent
请务必至少调度一个审查员。如果不确定调度哪个，就两个都调度——重复的审查结果总比遗漏问题要好。

Dispatch

调度审查员

Spawn agents in parallel when dispatching multiple. Each agent receives:
  1. The file paths to review
  2. The user's profile (language, provider, experience level)
  3. Which mode(s) to run (for chart-reviewer)
  4. Context: what the user said they built and what they're worried about
需要调度多个审查员时请并行创建实例。每个agent会收到以下信息：
  1. 待审查的文件路径
  2. 用户配置（语言、数据提供商、经验水平）
  3. 需要运行的模式（针对chart-reviewer）
  4. 上下文：用户说明的构建内容以及他们担心的问题

Data reviewer prompt template

数据审查员Prompt模板

Review the football data code in [FILE_PATHS].

The user is working with [PROVIDER] data in [LANGUAGE].
They built: [DESCRIPTION]
Their concern: [WHAT_THEY_SAID]

Follow the full review checklist in your agent prompt. Use search_docs to verify
provider-specific facts (coordinate systems, qualifier IDs, event types).
Review the football data code in [FILE_PATHS].

The user is working with [PROVIDER] data in [LANGUAGE].
They built: [DESCRIPTION]
Their concern: [WHAT_THEY_SAID]

Follow the full review checklist in your agent prompt. Use search_docs to verify
provider-specific facts (coordinate systems, qualifier IDs, event types).

Chart reviewer prompt template

图表审查员Prompt模板

Review the chart code in [FILE_PATHS].

Mode(s): [Code Review / Visual Inspection / Interactive Edge Cases]
The user is building: [DESCRIPTION]
Their concern: [WHAT_THEY_SAID]
Stack: [LANGUAGE + LIBRARIES from profile]
[If visual inspection: URL or instructions to render]

Load skills/brainstorm/references/chart-canon.md for convention checking.
Review the chart code in [FILE_PATHS].

Mode(s): [Code Review / Visual Inspection / Interactive Edge Cases]
The user is building: [DESCRIPTION]
Their concern: [WHAT_THEY_SAID]
Stack: [LANGUAGE + LIBRARIES from profile]
[If visual inspection: URL or instructions to render]

Load skills/brainstorm/references/chart-canon.md for convention checking.

Synthesise findings

整合审查结果

After both agents report back:
  1. Deduplicate — if both flag the same issue (e.g., wrong coordinate system), merge into one finding
  2. Sort by severity — Critical first, then Warning, then Info
  3. Group logically — Data issues, then Rendering issues, then Convention issues, then Edge cases
  4. Present concisely — table format with severity, location, issue, fix
两个agent都返回结果后：
  1. 去重——如果两者都标记了同一个问题（比如坐标系错误），合并为一条结果
  2. 按严重程度排序——严重优先级最高，然后是警告，最后是提示
  3. 按逻辑分组——先数据问题，然后渲染问题，然后规范问题，最后边界场景问题
  4. 简洁展示——用表格形式呈现，包含严重程度、位置、问题、修复方案

When to suggest visual inspection

何时建议进行视觉检查

If the chart-reviewer's code review finds potential rendering issues but can't confirm without seeing the output, suggest:
"The code review found [N] potential rendering issues. Want me to visually inspect the chart? I'll need a URL or you can run it locally."
Don't require visual inspection — many users can't easily serve their chart locally. Code review alone catches most issues.
如果chart-reviewer的代码审查发现了潜在的渲染问题，但看不到输出无法确认的话，建议：
"代码审查发现了[N]个潜在渲染问题。需要我对图表进行视觉检查吗？我需要一个URL，或者你可以在本地运行它。"
不要强制要求视觉检查——很多用户无法轻松在本地部署图表。仅代码审查就能发现绝大多数问题。

After review

审查结束后

If findings are found:
  • Ask the user which ones to fix
  • For Critical issues, offer to fix them directly
  • For Warning/Info, explain the trade-off and let them decide
If no findings:
  • Say so clearly. Don't invent issues to justify the review.
  • Optionally mention what was checked so the user knows the review was thorough.
如果发现了问题：
  • 询问用户想要修复哪些问题
  • 对于严重问题，主动提出直接修复
  • 对于警告/提示级别的问题，解释取舍方案，让用户自行决定
如果没有发现问题：
  • 明确告知用户，不要为了证明审查完成而编造问题
  • 可以选择性地提及检查了哪些内容，让用户知道审查是全面的。