code-debugging

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Code Debugging

代码调试

Systematically debug experiment code with structured error categorization and fix strategies.

通过结构化错误分类和修复策略，系统性地调试实验代码。

Input

输入

```
$0
```
— Error message, stderr output, or code file with issues
```
$1
```
— Optional: the code that produced the error

```
$0
```
— 错误信息、stderr输出或存在问题的代码文件
```
$1
```
— 可选：产生错误的对应代码

References

参考资料

Debug patterns and state machine:

~/.claude/skills/code-debugging/references/debug-patterns.md

调试模式与状态机：

~/.claude/skills/code-debugging/references/debug-patterns.md

Workflow

工作流程

Step 1: Categorize the Error

步骤1：错误分类

Category	Examples	Severity
SyntaxError	Invalid syntax, indentation	Low
ImportError	Missing module, wrong name	Low
RuntimeError	Division by zero, shape mismatch	Medium
TimeoutError	Infinite loop, too slow	Medium
OutputError	Missing files, wrong format	Medium
LogicError	Wrong results, 0% accuracy	High

分类	示例	严重程度
SyntaxError	语法无效、缩进错误	低
ImportError	缺失模块、名称错误	低
RuntimeError	除零错误、形状不匹配	中
TimeoutError	死循环、运行过慢	中
OutputError	文件缺失、格式错误	中
LogicError	结果错误、准确率0%	高

Step 2: Analyze Root Cause

步骤2：分析根本原因

Read the error traceback (last 1500 chars if truncated)
Identify the exact line and variable causing the error
Check for common patterns:
- Device mismatch (CPU vs GPU tensors)
- Shape mismatch in matrix operations
- Missing data normalization
- Off-by-one errors in indexing
- Incorrect loss function for task type

读取错误回溯信息（若被截断则取最后1500个字符）
定位导致错误的精确代码行和变量
检查常见错误模式：
- 设备不匹配（CPU 与 GPU 张量）
- 矩阵运算中的形状不匹配
- 缺失数据归一化
- 索引差一错误
- 任务类型对应的损失函数使用错误

Step 3: Apply Fix Strategy

步骤3：应用修复策略

For syntax/import errors: Direct fix, single attempt For runtime errors: Fix and rerun, up to 4 retries For logic errors: Reflect on approach, consider alternative methods For timeout: Reduce dataset size, optimize bottleneck, add early stopping

语法/导入错误：直接修复，单次尝试 运行时错误：修复后重新运行，最多重试4次 逻辑错误：复盘实现思路，考虑替代方案 超时问题：减小数据集规模、优化性能瓶颈、添加早停机制

Step 4: Reflect and Prevent

步骤4：复盘与预防

After fixing:

Explain why the error occurred
Identify which lines caused it
Describe the fix line-by-line
Note patterns to avoid in future code

修复完成后：

解释错误发生的原因
定位导致错误的代码行
逐行说明修复内容
记录后续代码需要规避的模式

Fix Strategy State Machine

修复策略状态机

Stage 0 (first attempt) → repost code as fresh
Stage 1 (second attempt) → repost or leave depending on severity
Stage 2 (third attempt) → regenerate from scratch if still failing

Stage 0 (first attempt) → repost code as fresh
Stage 1 (second attempt) → repost or leave depending on severity
Stage 2 (third attempt) → regenerate from scratch if still failing

Rules

规则

Prefer minimal targeted edits over full rewrites
Maximum 4-5 fix attempts before changing approach
Always truncate long error outputs to last 1500 characters
After fixing, verify the fix doesn't introduce new errors
Keep error history to avoid repeating the same mistakes
If 0% accuracy: check accuracy calculation first, then check data pipeline

优先采用最小范围的针对性修改，而非全量重写
调整方案前最多进行4-5次修复尝试
长错误输出始终截断至最后1500个字符
修复完成后验证不会引入新错误
留存错误历史，避免重复犯相同错误
若出现准确率0%的情况：先检查准确率计算逻辑，再检查数据管道