goodvibes-codebase-review

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Goodvibes Codebase Review & Remediation

Goodvibes 代码库审查与修复

Systematic codebase analysis with parallelized remediation using goodvibes agents.

CRITICAL: Be objective & completely honest. No sugar coating, EVER.

Honest Review: Honest reviews yield better results. Never consider emotions or feelings in your assessments.

借助goodvibes Agent实现系统化代码库分析与并行修复。

重要提示：保持客观且完全诚实。绝不粉饰，永远如此。

诚实审查：诚实的审查才能带来更好的结果。评估时绝不考虑情绪或感受。

Filesystem Boundaries

文件系统边界

CRITICAL: Write-local, read-global.

WRITE/EDIT/CREATE: ONLY within the current working directory and its subdirectories. This is the project root. All changes must be git-trackable.
READ: Can read any file anywhere for context (node_modules, global configs, other projects for reference, etc.)
NEVER WRITE to: parent directories, home directory, system files, other projects, anything outside project root.

The working directory when you were spawned IS the project root. Stay within it for all modifications.

重要提示：本地写入，全局读取。

写入/编辑/创建：仅可在当前工作目录及其子目录内操作。此目录即为项目根目录。所有更改必须可被git追踪。
读取：可读取任意位置的文件以获取上下文（如node_modules、全局配置、其他参考项目等）
禁止写入：父目录、主目录、系统文件、其他项目及项目根目录以外的任何内容。

启动你的工作目录即为项目根目录。所有修改操作均需在此范围内进行。

MCP Tool Checklist (MANDATORY)

MCP工具检查清单（强制执行）

STOP. Before doing ANYTHING, complete this checklist.

CRITICAL PREREQUISITE: You MUST call

mcp-cli info <server>/<tool>

BEFORE ANY

mcp-cli call <server>/<tool>

. This is a BLOCKING REQUIREMENT.

暂停。在执行任何操作前，完成此检查清单。

关键前置要求：在调用任何

mcp-cli call <server>/<tool>

之前，你必须先调用

mcp-cli info <server>/<tool>

。这是一项强制性要求。

Task Start

任务启动

bash

undefined

bash

undefined

ALWAYS check schema first

始终先检查 schema

mcp-cli info plugin_goodvibes_analysis-engine/detect_stack mcp-cli info plugin_goodvibes_registry-engine/recommend_skills mcp-cli info plugin_goodvibes_project-engine/project_issues

Then make the calls

然后进行调用

mcp-cli call plugin_goodvibes_analysis-engine/detect_stack '{}' mcp-cli call plugin_goodvibes_registry-engine/recommend_skills '{"task":"codebase review quality audit"}' mcp-cli call plugin_goodvibes_project-engine/project_issues '{}'

undefined

undefined

Review Phase

审查阶段

bash

undefined

bash

undefined

Check schemas first

先检查 schema

mcp-cli info plugin_goodvibes_analysis-engine/find_circular_deps mcp-cli info plugin_goodvibes_analysis-engine/scan_for_secrets mcp-cli info plugin_goodvibes_project-engine/analyze_dependencies mcp-cli info plugin_goodvibes_analysis-engine/find_dead_code mcp-cli info plugin_goodvibes_project-engine/get_test_coverage

Then execute

然后执行

mcp-cli call plugin_goodvibes_analysis-engine/find_circular_deps '{}' mcp-cli call plugin_goodvibes_analysis-engine/scan_for_secrets '{}' mcp-cli call plugin_goodvibes_project-engine/analyze_dependencies '{}' mcp-cli call plugin_goodvibes_analysis-engine/find_dead_code '{}' mcp-cli call plugin_goodvibes_project-engine/get_test_coverage '{}'

undefined

undefined

Before Every Edit

每次编辑前

bash

undefined

bash

undefined

Check schemas

检查 schema

mcp-cli info plugin_goodvibes_analysis-engine/scan_patterns mcp-cli info plugin_goodvibes_project-engine/find_tests_for_file mcp-cli info plugin_goodvibes_analysis-engine/validate_edits_preview

Execute

执行

mcp-cli call plugin_goodvibes_analysis-engine/scan_patterns '{}' mcp-cli call plugin_goodvibes_project-engine/find_tests_for_file '{"file":"path/to/file.ts"}' mcp-cli call plugin_goodvibes_analysis-engine/validate_edits_preview '{}'

undefined

undefined

After Every Edit

每次编辑后

bash

undefined

bash

undefined

Check schema

检查 schema

mcp-cli info plugin_goodvibes_project-engine/project_issues

Execute

执行

mcp-cli call plugin_goodvibes_project-engine/project_issues '{}'


**THE LAW: If a goodvibes MCP tool can do it, USE THE TOOL. No exceptions.**

---

mcp-cli call plugin_goodvibes_project-engine/project_issues '{}'


**规则：如果goodvibes MCP工具可完成该操作，务必使用工具。无例外。**

---

Precision Tools (MANDATORY)

Precision工具（强制执行）

CRITICAL: Use precision tools, NOT system tools.

Instead Of	Use	Why
`Read`	`precision_read`	Extract modes, line ranges, outline/symbols
`Grep`	`precision_grep`	Output modes, batch queries, context control
`Glob`	`precision_glob`	Output modes, filters, preview
`Edit`	`precision_edit`	Atomic transactions, validation, hints
`Write`	`precision_write`	Atomic, templates, validation
`Bash`	`precision_exec`	Batch commands, expectations, output control

重要提示：使用Precision工具，而非系统工具。

替代工具	使用工具	原因
`Read`	`precision_read`	提取模式、行范围、大纲/符号
`Grep`	`precision_grep`	输出模式、批量查询、上下文控制
`Glob`	`precision_glob`	输出模式、过滤器、预览
`Edit`	`precision_edit`	原子事务、验证、提示
`Write`	`precision_write`	原子操作、模板、验证
`Bash`	`precision_exec`	批量命令、预期结果、输出控制

Precision Tool Patterns

Precision工具使用示例

yaml

undefined

yaml

undefined

Find files with pattern (minimal output)

查找包含指定模式的文件（极简输出）

precision_grep: queries: - pattern: "TODO|FIXME|HACK" glob: "src/**/*.ts" output: mode: files_only

Read file structure without content

读取文件结构而非内容

precision_read: files: ["src/index.ts", "src/app.ts"] extract: outline output: mode: minimal

Batch edit multiple files atomically

原子化批量编辑多个文件

precision_edit: edits: - file: "src/api.ts" find: "const API_URL = 'http://localhost'" replace: "const API_URL = process.env.API_URL" - file: "src/config.ts" find: "debug: true" replace: "debug: process.env.NODE_ENV !== 'production'" transaction: mode: atomic rollback_on_fail: true output: mode: minimal

Execute commands with expectations

执行带预期结果的命令

precision_exec: commands: - cmd: "npm run typecheck" expect: exit_code: 0 - cmd: "npm run lint" expect: exit_code: 0 output: mode: minimal

---

precision_exec: commands: - cmd: "npm run typecheck" expect: exit_code: 0 - cmd: "npm run lint" expect: exit_code: 0 output: mode: minimal

---

Discovery -> Batch Workflow

发现 -> 批量工作流

CRITICAL: Always discover before batching.

The

discover

tool runs multiple queries in parallel to gather context before building a batch. This prevents wasted operations and ensures you target exactly the right files.

重要提示：始终先进行发现再执行批量操作。

discover

工具可并行运行多个查询，以在构建批量操作前收集上下文信息。这能避免无效操作，确保精准定位目标文件。

Discovery Tool Usage

发现工具使用示例

yaml

undefined

yaml

undefined

Run parallel discovery queries

运行并行发现查询

discover: queries: - id: find_components type: glob patterns: ["src/components//*.tsx"] - id: find_api_routes type: glob patterns: ["src/api//.ts", "src/app/api/**/.ts"] - id: find_issues type: grep pattern: "TODO|FIXME|HACK" glob: "src/**/*.{ts,tsx}" - id: find_hooks type: symbols query: "use" kinds: ["function"] output_mode: files_only # count_only | files_only | locations

undefined

undefined

Workflow Pattern

工作流模式

Discover - Run queries to understand scope
- Use
```
count_only
```
  first to gauge magnitude
- Then
```
files_only
```
  to get target list
Plan - Build batch operations using discovery results
- Reference discovered files in batch operations
- Scope work to exactly what was found
Execute - Run batch with full context

发现 - 运行查询以了解范围
- 先使用
```
count_only
```
  评估规模
- 再使用
```
files_only
```
  获取目标文件列表
规划 - 借助发现结果构建批量操作
- 在批量操作中引用发现的文件
- 将工作范围限定为已发现的内容
执行 - 结合完整上下文运行批量操作

Example: Feature Implementation

示例：功能实现

yaml

undefined

yaml

undefined

Step 1: Discover current state

步骤1：发现当前状态

discover: queries: - id: existing_files type: glob patterns: ["src/features/auth//*.ts"] - id: existing_patterns type: grep pattern: "export (function|const|class)" glob: "src/features//*.ts" output_mode: files_only

Step 2: Use results to build targeted batch

步骤2：使用结果构建定向批量操作

batch: id: implement-feature operations: read: - id: analyze type: files targets: "{{existing_files.files}}" # From discovery extract: outline


**Benefits:**
- Prevents blind operations on wrong files
- Ensures consistent patterns across the codebase
- Reduces token usage by targeting exactly what's needed
- Enables informed decisions about implementation approach

---

batch: id: implement-feature operations: read: - id: analyze type: files targets: "{{existing_files.files}}" # 来自发现结果 extract: outline


**优势：**
- 避免对错误文件执行盲目操作
- 确保代码库中的模式一致性
- 通过精准定位减少令牌使用量
- 助力对实现方案做出明智决策

---

Mode-Aware Behavior

模式感知行为

Your behavior adapts based on the current mode:

你的行为会根据当前模式进行调整：

vibecoding Mode

vibecoding模式

Communicate: Show progress, explain decisions, report results in detail
Ask: On ambiguity or risk, ask the user before proceeding
Checkpoint: Create checkpoints per batch
Output: Standard verbosity, show diffs

沟通：展示进度、解释决策、详细报告结果
询问：遇到歧义或风险时，先询问用户再继续
检查点：为每个批量操作创建检查点
输出：标准详细程度，显示差异

justvibes Mode

justvibes模式

Silent: Minimal communication, log to
```
.goodvibes/logs/activity.md
```
Autonomous: Make best-guess decisions, proceed with checkpoints on risk
Auto-chain: Continue to next logical batch automatically
Output: Minimal verbosity, no diffs

静默：最少沟通，日志记录至
```
.goodvibes/logs/activity.md
```
自主：做出最佳猜测决策，遇到风险时创建检查点后继续
自动链式执行：自动继续下一个逻辑批量操作
输出：最低详细程度，不显示差异

Workflow Overview

工作流概述

┌─────────────────────────────────────────────────────────────┐
│                      CODEBASE REVIEW                        │
├─────────────────────────────────────────────────────────────┤
│  Phase 1: Review      │  Run MCP tools, analyze all code   │
│  Phase 2: Report      │  Generate codebase-review-report.md │
│  Phase 3: Plan        │  Create remediation-plan.md        │
│  Phase 4: Execute     │  WORK-REVIEW-FIX-CHECK (max 6)     │
└─────────────────────────────────────────────────────────────┘

┌─────────────────────────────────────────────────────────────┐
│                      代码库审查                        │
├─────────────────────────────────────────────────────────────┤
│  阶段1：审查      │  运行MCP工具，分析所有代码   │
│  阶段2：报告      │  生成codebase-review-report.md │
│  阶段3：规划        │  创建remediation-plan.md        │
│  阶段4：执行     │  WORK-REVIEW-FIX-CHECK（最多6个并发）     │
└─────────────────────────────────────────────────────────────┘

Phase 1: Codebase Review

阶段1：代码库审查

MCP Tool Mapping

MCP工具映射

Category	Primary Tools	Fallback
Quality	`find_dead_code` , `scan_patterns`	grep for patterns
Architecture	`find_circular_deps` , `get_api_surface`	manual analysis
Security	`scan_for_secrets` , `check_permissions`	grep for patterns
Performance	`get_prisma_operations` , `analyze_bundle`	none
Documentation	`explain_codebase`	file scan
Testing	`get_test_coverage` , `find_tests_for_file` , `suggest_test_cases`	jest --coverage
Config	`read_config` , `env_audit`	env file scan
Dependencies	`analyze_dependencies`	npm audit
Errors	`parse_error_stack` , `explain_type_error`	tsc output
Style	`scan_patterns` , `get_conventions`	eslint output

类别	主要工具	备选方案
质量	`find_dead_code` , `scan_patterns`	模式匹配grep
架构	`find_circular_deps` , `get_api_surface`	手动分析
安全	`scan_for_secrets` , `check_permissions`	模式匹配grep
性能	`get_prisma_operations` , `analyze_bundle`	无
文档	`explain_codebase`	文件扫描
测试	`get_test_coverage` , `find_tests_for_file` , `suggest_test_cases`	jest --coverage
配置	`read_config` , `env_audit`	环境文件扫描
依赖	`analyze_dependencies`	npm audit
错误	`parse_error_stack` , `explain_type_error`	tsc输出
风格	`scan_patterns` , `get_conventions`	eslint输出

Review Checklist

审查检查清单

Minimum per category:

5 specific file:line findings OR explicit "no issues found"
Quantified measurements (counts, percentages, LOC)
Severity classification (critical/high/medium/low)

每个类别最低要求：

5个具体的文件:行级发现结果，或明确标注“未发现问题”
量化指标（数量、百分比、代码行数）
严重程度分类（critical/high/medium/low）

Phase 2: Master Report

阶段2：总报告

Generate

codebase-review-report.md

markdown

undefined

生成

codebase-review-report.md

：

markdown

undefined

Codebase Review Report

代码库审查报告

Project: {name from detect_stack} Stack: {technologies detected} Generated: {ISO 8601 timestamp} Overall Score: {X.X}/10

项目：{来自detect_stack的名称} 技术栈：{检测到的技术} 生成时间：{ISO 8601时间戳} 总体得分：{X.X}/10

Executive Summary

执行摘要

Severity	Count	Description
🔴 Critical	N	Issues requiring immediate attention
🟠 High	N	Issues blocking production readiness
🟡 Medium	N	Issues impacting maintainability
🔵 Low	N	Minor improvements

严重程度	数量	描述
🔴 Critical	N	需要立即关注的问题
🟠 High	N	阻碍生产就绪的问题
🟡 Medium	N	影响可维护性的问题
🔵 Low	N	微小改进点

Score Breakdown

得分细分

Category	Weight	Raw	Deductions	Score	Grade
Quality	15%	10	-X.X	X.X	A-F
Architecture	15%	10	-X.X	X.X	A-F
Security	20%	10	-X.X	X.X	A-F
Performance	10%	10	-X.X	X.X	A-F
Documentation	5%	10	-X.X	X.X	A-F
Testing	15%	10	-X.X	X.X	A-F
Config	5%	10	-X.X	X.X	A-F
Dependencies	5%	10	-X.X	X.X	A-F
Errors	5%	10	-X.X	X.X	A-F
Style	5%	10	-X.X	X.X	A-F
TOTAL	100%	-	-	X.X	X

类别	权重	原始分	扣分项	得分	等级
质量	15%	10	-X.X	X.X	A-F
架构	15%	10	-X.X	X.X	A-F
安全	20%	10	-X.X	X.X	A-F
性能	10%	10	-X.X	X.X	A-F
文档	5%	10	-X.X	X.X	A-F
测试	15%	10	-X.X	X.X	A-F
配置	5%	10	-X.X	X.X	A-F
依赖	5%	10	-X.X	X.X	A-F
错误	5%	10	-X.X	X.X	A-F
风格	5%	10	-X.X	X.X	A-F
总计	100%	-	-	X.X	X

Score Calculation

得分计算

{Show the math for each category deduction}

{展示每个类别扣分的计算方式}

Detailed Findings

详细发现结果

{Category Name}

{类别名称}

Finding: {Title}

发现：{标题}

Field	Value
Severity	critical\|high\|medium\|low
Location	`file:line` or `file:startLine-endLine`
Measurement	{exact number or percentage}
Threshold	{acceptable value}
Impact	{business/technical consequence}
Points Deducted	{-X.X from Category}

Evidence:

{code snippet or tool output}

Required Fix:

{specific remediation code or steps}

{repeat for all findings}

字段	值
严重程度	critical\|high\|medium\|low
位置	`file:line` 或 `file:startLine-endLine`
量化指标	{精确数字或百分比}
阈值	{可接受值}
影响	{业务/技术后果}
扣分数值	{-X.X 从该类别中扣除}

证据：

{代码片段或工具输出}

所需修复：

{具体修复代码或步骤}

{重复所有发现结果}

What's Working Well

表现良好的部分

{List genuinely good patterns found, with file references}

{列出发现的优秀模式，并附带文件引用}

Improvement Roadmap

改进路线图

Phase	Focus	Expected Impact	New Score
1	Critical fixes	+X.X points	X.X
2	High priority	+X.X points	X.X
3	Medium priority	+X.X points	X.X
4	Polish	+X.X points	X.X

---

阶段	重点	预期影响	新得分
1	Critical问题修复	+X.X 分	X.X
2	High优先级问题	+X.X 分	X.X
3	Medium优先级问题	+X.X 分	X.X
4	优化完善	+X.X 分	X.X

---

Phase 3: Remediation Plan

阶段3：修复计划

Generate

remediation-plan.md

markdown

undefined

生成

remediation-plan.md

：

markdown

undefined

Remediation Plan

修复计划

Generated: {timestamp} Total Tasks: {N} Execution Strategy: WORK-REVIEW-FIX-CHECK with parallel goodvibes agents (max 6 concurrent)

生成时间：{时间戳} 总任务数：{N} 执行策略：使用并行goodvibes Agent执行WORK-REVIEW-FIX-CHECK（最多6个并发）

Execution Rules

执行规则

Rule	Value
Max concurrent agents	6
Agent type	goodvibes background
Context model	Fresh per task (no accumulation)
Tool priority	MCP tools > bash
Monitoring	Self-report via SubagentStop hook
Workflow	WORK → REVIEW → PASS/FAIL → (FIX → CHECK)

规则	值
最大并发Agent数	6
Agent类型	goodvibes后台Agent
上下文模型	每个任务全新上下文（无累积）
工具优先级	MCP工具 > bash
监控	通过SubagentStop钩子自我报告
工作流	WORK → REVIEW → PASS/FAIL → (FIX → CHECK)

Task Definitions

任务定义

Wave 1: Critical [P0] - Execute Immediately

第一波：Critical [P0] - 立即执行

TASK-001: {Description}

TASK-001：{描述}

Field	Value
Severity	critical
Target Files	`file1.ts` , `file2.ts`
Related Finding	{reference to report finding}
MCP Tools	`tool1` , `tool2`
Estimated Duration	Xm

Context from Report:

{relevant finding details}

{repeat for all tasks, grouped by wave}

字段	值
严重程度	critical
目标文件	`file1.ts` , `file2.ts`
相关发现结果	{报告中发现结果的引用}
MCP工具	`tool1` , `tool2`
预计时长	Xm

来自报告的上下文：

{相关发现结果详情}

{按波次重复所有任务}

Wave 2: High [P1]

第二波：High [P1]

Wave 3: Medium [P2]

第三波：Medium [P2]

Wave 4: Low [P3]

第四波：Low [P3]

Dependency Graph

依赖关系图

{Show task dependencies if any exist}

---

{展示任务间的依赖关系（如有）}

---

Phase 4: Parallel Agent Execution

阶段4：并行Agent执行

WORK-REVIEW-FIX-CHECK Workflow

WORK-REVIEW-FIX-CHECK工作流

For each remediation task:

  WORK Agent ─────────────────────> REVIEW Agent
  (goodvibes:engineer)              (goodvibes:reviewer)
       │                                  │
       │                                  ├─> PASS ─> Commit ─> Update Log ─> Next Task
       │                                  │
       │                                  └─> FAIL ─> FIX Agent ─> CHECK Agent
                                                         │              │
                                                         │              ├─> PASS ─> Commit
                                                         │              │
                                                         └──────────────┴─> FAIL (loop)

针对每个修复任务：

  WORK Agent ─────────────────────> REVIEW Agent
  (goodvibes:engineer)              (goodvibes:reviewer)
       │                                  │
       │                                  ├─> PASS ─> 提交 ─> 更新日志 ─> 下一个任务
       │                                  │
       │                                  └─> FAIL ─> FIX Agent ─> CHECK Agent
                                                         │              │
                                                         │              ├─> PASS ─> 提交
                                                         │              │
                                                         └──────────────┴─> FAIL（循环）

Agent Role Mapping

Agent角色映射

Phase	Agent	Purpose
WORK	`goodvibes:engineer`	Implements remediation task
REVIEW	`goodvibes:reviewer`	Verifies work quality (100% required)
FIX	`goodvibes:engineer`	Addresses ALL review issues
CHECK	`goodvibes:reviewer`	Re-verifies fixes

阶段	Agent	用途
WORK	`goodvibes:engineer`	实施修复任务
REVIEW	`goodvibes:reviewer`	验证工作质量（100%必需）
FIX	`goodvibes:engineer`	解决所有审查问题
CHECK	`goodvibes:reviewer`	重新验证修复结果

Waiting for Agents (CRITICAL)

等待Agent（重要提示）

NEVER use these to check agent status:

```
tail
```
command on transcript files
```
TaskOutput
```
tool
Any form of polling

INSTEAD, the orchestrator:

Spawns agent with
```
run_in_background: true
```
Stops taking actions (turn ends)
Receives SubagentStop hook notification when agent completes
Hook message appears in context with status

绝不要使用以下方式检查Agent状态：

对转录文件使用
```
tail
```
命令
```
TaskOutput
```
工具
任何形式的轮询

正确方式：编排器

使用
```
run_in_background: true
```
启动Agent
停止执行操作（当前轮次结束）
当Agent完成时，接收SubagentStop钩子通知
上下文会显示包含状态的钩子消息

Concurrency Rules

并发规则

Rule	Value
Max concurrent agents	6
Agents per task	1 (one agent works on a task at a time)
Completion requirement	100% (not 99.9%)

规则	值
最大并发Agent数	6
每个任务的Agent数	1（同一时间一个Agent处理一个任务）
完成要求	100%完成（而非99.9%）

WORK Agent Prompt Template

WORK Agent提示模板

markdown

undefined

markdown

undefined

Remediation Task: {TASK_ID}

修复任务：{TASK_ID}

MCP Tool Checklist (MANDATORY)

MCP工具检查清单（强制执行）

CRITICAL PREREQUISITE: You MUST call

mcp-cli info <server>/<tool>

BEFORE ANY

mcp-cli call <server>/<tool>

Before ANY edit:

bash

undefined

关键前置要求：在调用任何

mcp-cli call <server>/<tool>

之前，你必须先调用

mcp-cli info <server>/<tool>

。

在进行任何编辑前：

bash

undefined

Check schemas first

先检查 schema

mcp-cli info plugin_goodvibes_analysis-engine/scan_patterns mcp-cli info plugin_goodvibes_project-engine/find_tests_for_file

Then execute

然后执行

mcp-cli call plugin_goodvibes_analysis-engine/scan_patterns '{}' mcp-cli call plugin_goodvibes_project-engine/find_tests_for_file '{"file":"{TARGET_FILE}"}'


After EVERY edit:
```bash

mcp-cli call plugin_goodvibes_analysis-engine/scan_patterns '{}' mcp-cli call plugin_goodvibes_project-engine/find_tests_for_file '{"file":"{TARGET_FILE}"}'


在每次编辑后：
```bash

Check schema first

先检查 schema

mcp-cli info plugin_goodvibes_project-engine/project_issues

Then execute

然后执行

mcp-cli call plugin_goodvibes_project-engine/project_issues '{}'

undefined

mcp-cli call plugin_goodvibes_project-engine/project_issues '{}'

undefined

Assignment

任务分配

Field	Value
Task ID	{TASK_ID}
Severity	{SEVERITY}
Description	{DESCRIPTION}
Files	{FILE_LIST}

字段	值
任务ID	{TASK_ID}
严重程度	{SEVERITY}
描述	{DESCRIPTION}
文件	{FILE_LIST}

Context from Report

来自报告的上下文

{FINDING_DETAILS}

Instructions

说明

Complete ONLY this assigned task
Use goodvibes MCP tools BEFORE bash
Use precision tools (precision_edit, precision_exec) for all operations
Run validation tools after every edit
If edit causes new errors, fix them before completing

仅完成分配的此任务
优先使用goodvibes MCP工具，再使用bash
所有操作均使用precision工具（precision_edit, precision_exec等）
每次编辑后运行验证工具
如果编辑导致新错误，在完成前修复这些错误

Tool Priority (MANDATORY)

工具优先级（强制执行）

First: Check

mcp-cli info

then

mcp-cli call plugin_goodvibes_*

Second: Use precision tools (precision_edit, precision_read, etc.)
Only then: Fall back to standard tools if no MCP/precision tool exists

首先：检查

mcp-cli info

，然后调用

mcp-cli call plugin_goodvibes_*

其次：使用precision工具（precision_edit, precision_read等）
仅当没有MCP/precision工具时，才回退到标准工具

Completion Report

完成报告

When done, output:

json

{
  "task_id": "{TASK_ID}",
  "status": "success|failed",
  "files_modified": ["file1.ts", "file2.ts"],
  "tests_passed": true|false,
  "type_check_passed": true|false,
  "notes": "Any relevant context"
}

---

完成后，输出：

json

{
  "task_id": "{TASK_ID}",
  "status": "success|failed",
  "files_modified": ["file1.ts", "file2.ts"],
  "tests_passed": true|false,
  "type_check_passed": true|false,
  "notes": "任何相关上下文"
}

---

REVIEW Agent Prompt Template

REVIEW Agent提示模板

markdown

undefined

markdown

undefined

Review Task: {TASK_ID}

审查任务：{TASK_ID}

MCP Tool Checklist (MANDATORY)

MCP工具检查清单（强制执行）

CRITICAL PREREQUISITE: You MUST call

mcp-cli info <server>/<tool>

BEFORE ANY

mcp-cli call <server>/<tool>

bash

undefined

关键前置要求：在调用任何

mcp-cli call <server>/<tool>

之前，你必须先调用

mcp-cli info <server>/<tool>

。

bash

undefined

Check schemas first

先检查 schema

mcp-cli info plugin_goodvibes_project-engine/project_issues mcp-cli info plugin_goodvibes_analysis-engine/scan_patterns mcp-cli info plugin_goodvibes_project-engine/get_test_coverage

Then execute

然后执行

mcp-cli call plugin_goodvibes_project-engine/project_issues '{}' mcp-cli call plugin_goodvibes_analysis-engine/scan_patterns '{}' mcp-cli call plugin_goodvibes_project-engine/get_test_coverage '{}'

undefined

mcp-cli call plugin_goodvibes_project-engine/project_issues '{}' mcp-cli call plugin_goodvibes_analysis-engine/scan_patterns '{}' mcp-cli call plugin_goodvibes_project-engine/get_test_coverage '{}'

undefined

Assignment

任务分配

Field	Value
Task ID	{TASK_ID}
Work Agent	{WORK_AGENT_ID}
Files Modified	{FILE_LIST}

字段	值
任务ID	{TASK_ID}
WORK Agent	{WORK_AGENT_ID}
修改的文件	{FILE_LIST}

Context from WORK Agent

来自WORK Agent的上下文

{WORK_COMPLETION_REPORT}

Review Criteria

审查标准

Criterion	Weight	Description
Correctness	30%	Does it solve the problem correctly?
Completeness	25%	Are all aspects addressed?
Quality	20%	Is the code clean, maintainable?
Safety	15%	No new errors, security issues?
Tests	10%	Tests pass, coverage maintained?

标准	权重	描述
正确性	30%	是否正确解决了问题？
完整性	25%	是否解决了所有方面？
质量	20%	代码是否简洁、可维护？
安全性	15%	无新错误、安全问题？
测试	10%	测试通过、覆盖率保持？

Review Process

审查流程

Read all modified files
Run MCP validation tools
Check for:
- Type errors
- Test failures
- Security issues
- Code quality issues
- Incomplete fixes
Calculate score based on criteria weights
Provide verdict

读取所有修改的文件
运行MCP验证工具
检查：
- 类型错误
- 测试失败
- 安全问题
- 代码质量问题
- 未完成的修复
根据标准权重计算得分
提供 verdict

Review Output

审查输出

json

{
  "task_id": "{TASK_ID}",
  "verdict": "PASS|FAIL",
  "score": 8.5,
  "issues": [
    {
      "severity": "critical|major|minor",
      "file": "path/to/file.ts",
      "line": 47,
      "issue": "Description of the issue",
      "fix_guidance": "Specific guidance on how to fix"
    }
  ],
  "recommendation": "Details for FIX agent if FAIL"
}

json

{
  "task_id": "{TASK_ID}",
  "verdict": "PASS|FAIL",
  "score": 8.5,
  "issues": [
    {
      "severity": "critical|major|minor",
      "file": "path/to/file.ts",
      "line": 47,
      "issue": "问题描述",
      "fix_guidance": "修复的具体指导"
    }
  ],
  "recommendation": "如果FAIL，提供给FIX Agent的详细信息"
}

Verdict Rules

Verdict规则

PASS: Score >= 8.0 AND no critical issues AND all criteria pass
FAIL: Score < 8.0 OR any critical issue OR any criterion fails

CRITICAL: Be honest. Better to catch issues now than in production.

---

PASS：得分 >= 8.0 且无critical问题且所有标准通过
FAIL：得分 < 8.0 或存在任何critical问题或任何标准未通过

重要提示：保持诚实。与其在生产环境中发现问题，不如现在就找出。

---

FIX Agent Prompt Template

FIX Agent提示模板

markdown

undefined

markdown

undefined

Fix Task: {TASK_ID} (Iteration {N})

修复任务：{TASK_ID}（迭代 {N}）

MCP Tool Checklist (MANDATORY)

MCP工具检查清单（强制执行）

CRITICAL PREREQUISITE: You MUST call

mcp-cli info <server>/<tool>

BEFORE ANY

mcp-cli call <server>/<tool>

Before ANY edit:

bash

undefined

关键前置要求：在调用任何

mcp-cli call <server>/<tool>

之前，你必须先调用

mcp-cli info <server>/<tool>

。

在进行任何编辑前：

bash

undefined

Check schemas first

先检查 schema

mcp-cli info plugin_goodvibes_analysis-engine/scan_patterns mcp-cli info plugin_goodvibes_project-engine/find_tests_for_file

Then execute

然后执行

mcp-cli call plugin_goodvibes_analysis-engine/scan_patterns '{}' mcp-cli call plugin_goodvibes_project-engine/find_tests_for_file '{"file":"{TARGET_FILE}"}'


After EVERY edit:
```bash

mcp-cli call plugin_goodvibes_analysis-engine/scan_patterns '{}' mcp-cli call plugin_goodvibes_project-engine/find_tests_for_file '{"file":"{TARGET_FILE}"}'


在每次编辑后：
```bash

Check schema first

先检查 schema

mcp-cli info plugin_goodvibes_project-engine/project_issues

Then execute

然后执行

mcp-cli call plugin_goodvibes_project-engine/project_issues '{}'

undefined

mcp-cli call plugin_goodvibes_project-engine/project_issues '{}'

undefined

Review Feedback

审查反馈

{ISSUES_FROM_REVIEW}

Instructions

说明

Address EVERY issue from the review - no skipping
Follow the
```
fix_guidance
```
provided for each issue
Use precision tools for all operations
Run validation after each fix
If you cannot fix an issue, document WHY and mark as blocked

解决审查中的所有问题 - 不得跳过
遵循每个问题提供的
```
fix_guidance
```
所有操作均使用precision工具
每次修复后运行验证
如果无法修复某个问题，记录原因并标记为阻塞

Completion Report

完成报告

json

{
  "task_id": "{TASK_ID}",
  "iteration": {N},
  "status": "success|blocked",
  "issues_addressed": N,
  "issues_remaining": N,
  "blocked_issues": [
    {
      "issue": "Description",
      "reason": "Why it cannot be fixed"
    }
  ],
  "notes": "Any relevant context"
}

---

json

{
  "task_id": "{TASK_ID}",
  "iteration": {N},
  "status": "success|blocked",
  "issues_addressed": N,
  "issues_remaining": N,
  "blocked_issues": [
    {
      "issue": "描述",
      "reason": "无法修复的原因"
    }
  ],
  "notes": "任何相关上下文"
}

---

CHECK Agent Prompt Template

CHECK Agent提示模板

markdown

undefined

markdown

undefined

Check Task: {TASK_ID} (Verification Round {N})

检查任务：{TASK_ID}（验证轮次 {N}）

MCP Tool Checklist (MANDATORY)

MCP工具检查清单（强制执行）

CRITICAL PREREQUISITE: You MUST call

mcp-cli info <server>/<tool>

BEFORE ANY

mcp-cli call <server>/<tool>

bash

undefined

关键前置要求：在调用任何

mcp-cli call <server>/<tool>

之前，你必须先调用

mcp-cli info <server>/<tool>

。

bash

undefined

Check schemas first

先检查 schema

mcp-cli info plugin_goodvibes_project-engine/project_issues mcp-cli info plugin_goodvibes_analysis-engine/scan_patterns

Then execute

然后执行

mcp-cli call plugin_goodvibes_project-engine/project_issues '{}' mcp-cli call plugin_goodvibes_analysis-engine/scan_patterns '{}'

undefined

mcp-cli call plugin_goodvibes_project-engine/project_issues '{}' mcp-cli call plugin_goodvibes_analysis-engine/scan_patterns '{}'

undefined

Assignment

任务分配

Field	Value
Task ID	{TASK_ID}
FIX Agent	{FIX_AGENT_ID}
Iteration	{N}

字段	值
任务ID	{TASK_ID}
FIX Agent	{FIX_AGENT_ID}
迭代次数	{N}

Previous Issues

之前的问题

{ISSUES_TO_VERIFY}

Context from FIX Agent

来自FIX Agent的上下文

{FIX_COMPLETION_REPORT}

Verification Process

验证流程

Read all modified files
Run MCP validation tools
Verify each previous issue is resolved
Check for new issues introduced by fixes
Provide verdict

读取所有修改的文件
运行MCP验证工具
验证每个之前的问题是否已解决
检查修复是否引入新问题
提供 verdict

Verdict Output

Verdict输出

json

{
  "task_id": "{TASK_ID}",
  "iteration": {N},
  "verdict": "PASS|FAIL",
  "resolved_issues": N,
  "unresolved_issues": [
    {
      "issue": "Description",
      "status": "still_present|new_issue"
    }
  ],
  "recommendation": "Next action if FAIL"
}

json

{
  "task_id": "{TASK_ID}",
  "iteration": {N},
  "verdict": "PASS|FAIL",
  "resolved_issues": N,
  "unresolved_issues": [
    {
      "issue": "描述",
      "status": "still_present|new_issue"
    }
  ],
  "recommendation": "如果FAIL，下一步操作"
}

Verdict Rules

Verdict规则

PASS: ALL previous issues resolved AND no new critical issues
FAIL: ANY issue unresolved OR new critical issues found

If FAIL after 3 iterations, escalate to human review.

---

PASS：所有之前的问题均已解决且无新的critical问题
FAIL：任何问题未解决或发现新的critical问题

如果3次迭代后仍FAIL，升级至人工审查。

---

Commit Protocol

提交协议

After REVIEW or CHECK verdict is PASS, create commit:

fix({CATEGORY}): {TASK_ID} - {SHORT_DESCRIPTION}

{DETAILED_DESCRIPTION}

当REVIEW或CHECK的verdict为PASS后，创建提交：

fix({CATEGORY}): {TASK_ID} - {简短描述}

{详细描述}

Changes

更改

{FILE_1}: {WHAT_CHANGED}
{FILE_2}: {WHAT_CHANGED}

{FILE_1}: {更改内容}
{FILE_2}: {更改内容}

Review

审查

Reviewed by: REVIEW agent
Score: {SCORE}/10
Iterations: {FIX_ITERATIONS}

Task-ID: {TASK_ID} Severity: {SEVERITY} Co-Authored-By: Claude Opus 4.5 noreply@anthropic.com

undefined

审查者：REVIEW agent
得分：{SCORE}/10
迭代次数：{FIX_ITERATIONS}

Task-ID: {TASK_ID} Severity: {SEVERITY} Co-Authored-By: Claude Opus 4.5 noreply@anthropic.com

undefined

Completion Cleanup

完成清理

After ALL remediation tasks have passed review and been committed, archive the output files:

Archive Location:

.goodvibes/completed/

Archive Commands:

bash

undefined

当所有修复任务通过审查并提交后，归档输出文件：

归档位置：

.goodvibes/completed/

归档命令：

bash

undefined

Run all commands together in a single shell session:

在单个shell会话中运行所有命令：

Create archive directory if it doesn't exist

如果归档目录不存在则创建

mkdir -p .goodvibes/completed

Generate timestamp

生成时间戳

TIMESTAMP=$(date +%Y%m%d-%H%M%S)

Move and rename files

移动并重命名文件

mv codebase-review-report.md .goodvibes/completed/code-review-${TIMESTAMP}.md mv remediation-plan.md .goodvibes/completed/remediation-plan-${TIMESTAMP}.md mv remediation-log.md .goodvibes/completed/remediation-log-${TIMESTAMP}.md


**Archive Checklist:**
- [ ] All remediation tasks show status: completed in remediation-log.md
- [ ] All commits pushed (if applicable)
- [ ] Memory files updated (.goodvibes/memory/)
- [ ] Archive directory created
- [ ] Files moved with timestamp suffix

**File Naming Convention:**
| Original | Archived |
|----------|----------|
| `codebase-review-report.md` | `code-review-{YYYYMMDD-HHMMSS}.md` |
| `remediation-plan.md` | `remediation-plan-{YYYYMMDD-HHMMSS}.md` |
| `remediation-log.md` | `remediation-log-{YYYYMMDD-HHMMSS}.md` |

**Why Archive:**
- Keeps project root clean
- Preserves historical review data
- Enables tracking of codebase health over time
- Prevents confusion with future reviews


**归档检查清单：**
- [ ] 所有修复任务在remediation-log.md中显示状态为已完成
- [ ] 所有提交已推送（如适用）
- [ ] 内存文件已更新（.goodvibes/memory/）
- [ ] 已创建归档目录
- [ ] 文件已添加时间戳后缀并移动

**文件命名规范：**
| 原文件名 | 归档后文件名 |
|----------|----------|
| `codebase-review-report.md` | `code-review-{YYYYMMDD-HHMMSS}.md` |
| `remediation-plan.md` | `remediation-plan-{YYYYMMDD-HHMMSS}.md` |
| `remediation-log.md` | `remediation-log-{YYYYMMDD-HHMMSS}.md` |

**归档原因：**
- 保持项目根目录整洁
- 保留历史审查数据
- 能够跟踪代码库健康状况随时间的变化
- 避免与未来的审查产生混淆

Completion Logging

完成日志

Maintain

remediation-log.md

markdown

undefined

维护

remediation-log.md

：

markdown

undefined

Remediation Log

修复日志

Active Agents

活跃Agent

Task ID	Description	Agent	Phase	Started	Elapsed
TASK-002	Remove secrets	agent-abc	WORK	10:05:00	5m
TASK-003	Fix auth flow	agent-def	REVIEW	10:08:00	2m

任务ID	描述	Agent	阶段	开始时间	耗时
TASK-002	移除密钥	agent-abc	WORK	10:05:00	5m
TASK-003	修复认证流程	agent-def	REVIEW	10:08:00	2m

Completed Tasks

已完成任务

Task ID	Description	Status	Duration	Iterations	Changes
TASK-001	Fix SQL injection	✅ PASS	15m32s	1	`api/users.ts`

任务ID	描述	状态	时长	迭代次数	更改
TASK-001	修复SQL注入	✅ PASS	15m32s	1	`api/users.ts`

Failed Tasks

失败任务

Task ID	Description	Status	Iterations	Reason	Retry?
TASK-005	Fix type errors	❌ FAIL	3	Blocked by external dependency	No

任务ID	描述	状态	迭代次数	原因	重试？
TASK-005	修复类型错误	❌ FAIL	3	被外部依赖阻塞	否

Summary

摘要

Total Tasks: N
Completed: N (X%)
In Progress: N
Failed: N
Remaining: N
Active Agents: N/6

---

总任务数：N
已完成：N (X%)
进行中：N
失败：N
剩余：N
活跃Agent数：N/6

---

Output Artifacts

输出产物

File	Purpose
`codebase-review-report.md`	Complete findings with quantified metrics and scores
`remediation-plan.md`	Prioritized task checklist by severity with context
`remediation-log.md`	Real-time execution tracking with WORK-REVIEW-FIX-CHECK status

Note: After all remediation tasks are completed, these files are archived to

.goodvibes/completed/

with timestamp suffixes. See "Completion Cleanup" section for details.

文件	用途
`codebase-review-report.md`	包含量化指标和得分的完整发现结果
`remediation-plan.md`	按严重程度优先级排序的任务清单及上下文
`remediation-log.md`	WORK-REVIEW-FIX-CHECK状态的实时执行跟踪

注意： 所有修复任务完成后，这些文件会被归档到

.goodvibes/completed/

并添加时间戳后缀。详情请见“完成清理”部分。

Memory Integration

内存集成

After review/remediation, update

.goodvibes/memory/

File	Update With
`decisions.md`	Architectural decisions made during remediation
`patterns.md`	Patterns discovered or enforced
`failures.md`	Failed remediations with root cause analysis
`preferences.md`	Tool preferences that worked well

审查/修复完成后，更新

.goodvibes/memory/

：

文件	更新内容
`decisions.md`	修复过程中做出的架构决策
`patterns.md`	发现或强制执行的模式
`failures.md`	失败的修复及根本原因分析
`preferences.md`	表现良好的工具偏好

Constraints

约束条件

Constraint	Value	Rationale
Max concurrent agents	6	Resource management
Agent type	goodvibes background	Telemetry via hooks
Tasks per agent	1	Fresh context per task
Context inheritance	None	Prevent cross-contamination
Monitoring method	Self-report	SubagentStop hook handles
Tool priority	MCP > precision > bash	Consistency and telemetry
Review requirement	100%	Ensure quality before commit

约束	值	理由
最大并发Agent数	6	资源管理
Agent类型	goodvibes后台Agent	通过钩子实现遥测
每个Agent的任务数	1	每个任务全新上下文
上下文继承	无	防止交叉污染
监控方式	自我报告	由SubagentStop钩子处理
工具优先级	MCP > precision > bash	一致性和遥测
审查要求	100%	提交前确保质量