iterative-refinement

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Iterative Refinement

迭代优化

Execute workflows iteratively with systematic validation, progress tracking, and intelligent termination.

通过系统化验证、进度跟踪和智能终止机制，以迭代方式执行工作流。

When to Use

适用场景

Use for tasks requiring iterative refinement:

Test-fix-validate cycles: Fix failures → retest → repeat until passing
Code quality improvement: Review → fix → review until standards met
Performance optimization: Profile → optimize → measure until targets achieved
Progressive enhancement: Iterative improvements until diminishing returns

Don't use for single-pass tasks, purely parallel work, or simple linear workflows.

适用于需要迭代优化的任务：

测试-修复-验证循环：修复问题 → 重新测试 → 重复直至全部通过
代码质量提升：评审 → 修复 → 评审直至符合标准
性能优化：分析 → 优化 → 度量直至达成目标
渐进式增强：持续迭代改进直至收益递减

不适用于单次完成的任务、纯并行工作或简单线性工作流。

Pre-Usage Research (Optional)

预使用调研（可选）

Before starting iterations, consider researching:

Current best practices for your validation tools (search "[tool] best practices 2025")
Known issues with your tech stack (search "[language] [tool] common issues")
Optimal configuration for your validators (search "[tool] configuration production")
Recent improvements or alternatives (search "[tool] vs alternatives 2025")

Benefits:

Better validators from the start
Avoid known issues
Use current best practices
Save iteration cycles

When to research first:

Unfamiliar validation tools
New tech stack
Complex quality criteria
High-stakes optimization

开始迭代前，可考虑调研：

验证工具的当前最佳实践（搜索 "[tool] best practices 2025"）
技术栈的已知问题（搜索 "[language] [tool] common issues"）
验证工具的最优配置（搜索 "[tool] configuration production"）
近期改进或替代方案（搜索 "[tool] vs alternatives 2025"）

调研收益：

从一开始就使用更优质的验证工具
规避已知问题
采用当前最佳实践
减少迭代次数

优先调研的场景：

不熟悉的验证工具
全新技术栈
复杂的质量标准
高风险优化任务

Core Loop Pattern

核心循环模式

Every iteration follows:

Execute action (fix, optimize, improve)
Validate result (test, measure, check)
Assess progress (compare to criteria)
Decide (continue or stop)

每次迭代遵循以下步骤：

执行操作（修复、优化、改进）
验证结果（测试、度量、检查）
评估进度（与标准对比）
决策（继续或终止）

Instructions

操作步骤

Step 1: Define Configuration

步骤1：定义配置

Establish before starting:

Success Criteria (specific and measurable):

Criterion 1: [Example: "All 50 tests passing"]
Criterion 2: [Example: "Zero linter warnings"]
Criterion 3: [Example: "Response time < 100ms"]

Loop Limits:

Max iterations: 5-15 (justify if >20)
Min iterations: (optional)

Termination Mode:

Fixed: Run exactly N iterations
Criteria: Stop when success criteria met
Convergence: Stop when improvements < threshold (e.g., <10% over 3 iterations)
Hybrid: Combine multiple conditions

开始前需确定：

成功标准（具体可衡量）：

标准1：[示例："全部50项测试通过"]
标准2：[示例："无代码检查工具警告"]
标准3：[示例："响应时间 < 100ms"]

循环限制：

最大迭代次数：5-15次（超过20次需说明理由）
最小迭代次数：（可选）

终止模式：

固定次数：恰好执行N次迭代
达标终止：满足成功标准时停止
收敛终止：连续3次迭代改进幅度<阈值（如<10%）时停止
混合模式：结合多种终止条件

Step 2: Execute Iteration

步骤2：执行迭代

For each iteration:

Take action - Apply fixes or implement changes
Run validator - Execute tests, linters, or measurements

Record progress:

Iteration N:
- Action: [what was done]
- Results: [metrics/outcomes]
- Issues remaining: [count/description]
- Decision: [Continue/Success/Stop]

Assess termination:
- All criteria met? → SUCCESS
- Improvement < threshold? → CONVERGED
- Reached max iterations? → STOP
- Otherwise → CONTINUE

每次迭代流程：

执行操作 - 修复问题或实施改进
运行验证 - 执行测试、代码检查或度量

记录进度：

第N次迭代:
- 操作：[执行的内容]
- 结果：[指标/产出]
- 剩余问题：[数量/描述]
- 决策：[继续/成功/终止]

评估终止条件：
- 满足所有标准？→ 成功
- 改进幅度<阈值？→ 已收敛
- 达到最大迭代次数？→ 终止
- 其他情况 → 继续

Step 3: Pass Context Between Iterations

步骤3：迭代间传递上下文

Each iteration needs:

Previous results
Current metrics
Remaining issues
Progress trend

This prevents repeating failed approaches.

每次迭代需获取：

上一次的结果
当前指标
剩余问题
进度趋势

避免重复已失败的方案。

Step 4: Handle Stuck States

步骤4：处理停滞状态

If no progress for 2-3 iterations:

Analyze why progress stopped
Try different approach
Consider manual intervention
Stop if truly stuck

若连续2-3次迭代无进展：

分析停滞原因
尝试不同方案
考虑人工干预
确认无法推进则终止

Step 5: Report Results

步骤5：结果报告

Loop Summary:
- Iterations: N
- Termination: [Success/Converged/Max/Stuck]
- Initial state: [metrics]
- Final state: [metrics]
- Improvement: [percentage/delta]
- Remaining issues: [list if any]

循环总结:
- 迭代次数：N
- 终止原因：[成功/收敛/达到最大次数/停滞]
- 初始状态：[指标]
- 最终状态：[指标]
- 改进幅度：[百分比/差值]
- 剩余问题：[列表（如有）]

Validation Best Practices

验证最佳实践

Make Validators Specific

验证标准需具体

Bad: "Check if code is better" Good: "Run linter and count warnings"

Bad: "See if it's faster" Good: "Run benchmark: average response time over 100 requests"

错误示例："检查代码是否更优" 正确示例："运行代码检查工具并统计警告数量"

错误示例："看看是否更快" 正确示例："运行基准测试：100次请求的平均响应时间"

Use Automated Validation

使用自动化验证

Prefer scripts/tools over manual inspection:

Test frameworks over reading test code
Linters over manual code review
Benchmarks over estimated performance
Coverage tools over counting tests

优先使用脚本/工具而非人工检查：

测试框架而非人工阅读测试代码
代码检查工具而非人工代码评审
基准测试而非估算性能
覆盖率工具而非人工统计测试数量

Capture Concrete Metrics

追踪具体指标

Track measurable progress:

Test pass rate: 42/50 → 48/50 → 50/50
Warning count: 23 → 8 → 2 → 0
Response time: 320ms → 180ms → 95ms → 48ms
Code coverage: 65% → 78% → 85% → 92%

记录可衡量的进度：

测试通过率：42/50 → 48/50 → 50/50
警告数量：23 → 8 → 2 → 0
响应时间：320ms → 180ms → 95ms → 48ms
代码覆盖率：65% → 78% → 85% → 92%

Examples

示例

Example 1: Test Fixing

示例1：测试修复

Task: Fix all failing tests

Configuration:

Success: 100% tests passing
Max iterations: 8

Execution:

I1: 42/50 → Fix 8 failures → Continue
I2: 48/50 → Fix 2 failures → Continue
I3: 50/50 → SUCCESS ✓

任务：修复所有失败的测试

配置：

成功标准：100%测试通过
最大迭代次数：8

执行过程：

I1: 42/50 → 修复8个失败项 → 继续
I2: 48/50 → 修复2个失败项 → 继续
I3: 50/50 → 成功 ✓

Example 2: Linter Cleanup

示例2：代码检查清理

Task: Remove all linter warnings

Configuration:

Success: 0 warnings
Max iterations: 5

Execution:

I1: 15 warnings → Fix → 6 warnings
I2: 6 warnings → Fix → 1 warning
I3: 1 warning → Fix → 0 warnings ✓

任务：清除所有代码检查警告

配置：

成功标准：0个警告
最大迭代次数：5

执行过程：

I1: 15个警告 → 修复 → 6个警告
I2: 6个警告 → 修复 → 1个警告
I3: 1个警告 → 修复 → 0个警告 ✓

Example 3: Performance Loop

示例3：性能优化循环

Task: Optimize response time

Configuration:

Success: <50ms OR converged
Max iterations: 15
Convergence: <10% over 3 iterations

Execution:

I1: 320ms → Optimize → 180ms (44%)
I2: 180ms → Optimize → 95ms (47%)
I3: 95ms → Optimize → 48ms (49%)
SUCCESS (target met)

任务：优化响应时间

配置：

成功标准：<50ms 或已收敛
最大迭代次数：15
收敛条件：连续3次迭代改进幅度<10%

执行过程：

I1: 320ms → 优化 → 180ms（提升44%）
I2: 180ms → 优化 → 95ms（提升47%）
I3: 95ms → 优化 → 48ms（提升49%）
成功（达标）

Example 4: Coverage Improvement

示例4：覆盖率提升

Task: Increase test coverage to 90%

Configuration:

Success: Coverage ≥ 90%
Max iterations: 12

Execution:

I1: 65% → Write tests → 72%
I2: 72% → Write tests → 81%
I3: 81% → Write tests → 88%
I4: 88% → Write tests → 91% ✓

任务：将测试覆盖率提升至90%

配置：

成功标准：覆盖率≥90%
最大迭代次数：12

执行过程：

I1: 65% → 编写测试 → 72%
I2: 72% → 编写测试 → 81%
I3: 81% → 编写测试 → 88%
I4: 88% → 编写测试 → 91% ✓

Language-Specific Tools

各语言专用工具

For validation tools and commands for your language:

Python: See tools/python.md
JavaScript/TypeScript: See tools/javascript.md
Rust: See tools/rust.md
Java: See tools/java.md
Go: See tools/go.md
C/C++: See tools/cpp.md
Ruby: See tools/ruby.md
PHP: See tools/php.md
C#/.NET: See tools/dotnet.md

各语言的验证工具及命令参考：

Python: See tools/python.md
JavaScript/TypeScript: See tools/javascript.md
Rust: See tools/rust.md
Java: See tools/java.md
Go: See tools/go.md
C/C++: See tools/cpp.md
Ruby: See tools/ruby.md
PHP: See tools/php.md
C#/.NET: See tools/dotnet.md

Advanced Usage

高级用法

For complex workflows, convergence detection, and advanced patterns: See patterns.md

复杂工作流、收敛检测及高级模式参考： See patterns.md

Best Practices

最佳实践

DO:

建议：

✓ Define clear, measurable success criteria ✓ Set reasonable max limits (5-15) ✓ Use automated validators ✓ Pass context between iterations ✓ Track concrete metrics ✓ Stop early when criteria met ✓ Detect convergence ✓ Document changes

✓ 定义清晰、可衡量的成功标准 ✓ 设置合理的最大迭代次数（5-15次） ✓ 使用自动化验证工具 ✓ 在迭代间传递上下文 ✓ 追踪具体指标 ✓ 达标后尽早终止 ✓ 检测收敛状态 ✓ 记录变更内容

DON'T:

禁忌：

✗ Use loops for single-pass tasks ✗ Set high limits (>20) without justification ✗ Skip validation between iterations ✗ Lose context between iterations ✗ Continue after success/convergence ✗ Ignore stuck signals ✗ Use vague criteria ✗ Miss early termination

✗ 对单次任务使用循环 ✗ 无理由设置高迭代次数（>20次） ✗ 跳过迭代间的验证步骤 ✗ 丢失迭代间的上下文 ✓ 达标/收敛后仍继续迭代 ✗ 忽略停滞信号 ✗ 使用模糊的标准 ✗ 错过提前终止时机