content-refiner

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Content Refiner (The Fixer)

内容优化工具（修复器）

Purpose

用途

POST-GATE TOOL. Transforms content that FAILED Gate 4 into passing content. Focuses on trimming verbosity and fixing continuity.

POST-GATE工具 将未通过Gate 4的内容转化为合格内容。重点在于精简冗余表述并修复内容连贯性。

When to Use

使用时机

Trigger: Gate 4 (Acceptance Auditor) returned
```
[FAIL]
```
.
Goal: Fix word count OR continuity issues (or both).
Key: Diagnose what failed BEFORE applying fixes.

触发条件：Gate 4（验收审核器）返回
```
[FAIL]
```
。
目标：修复字数问题或连贯性问题（或两者皆有）。
关键：在应用修复前先诊断失败原因。

CRITICAL: Pre-Refinement Diagnosis

重要提示：优化前诊断

DO NOT apply fixes blindly. Gate 4 fails for different reasons requiring different strategies.

请勿盲目应用修复。Gate 4不通过的原因不同，需要采用不同的策略。

Step 0: Identify What Failed (Mandatory)

步骤0：确定失败原因（必填）

Ask the user OR examine the Gate 4 failure message:

Failure Type	Question	Action
Word Count	"Is the lesson over the target (typically 1500 words)?"	Calculate exact % to cut
Continuity	"Does the opening reference the previous lesson?"	Rewrite opening only
Both	"Word count AND continuity broken?"	Two-phase approach

DIAGNOSIS EXAMPLES:

Example 1: Word Count Only

Content: 1950 words, Target: 1500
Excess: 450 words
% to cut: (450 / 1950) × 100 = 23%
→ CUT EXACTLY 23%, not generic 15-20%

Example 2: Continuity Only

Opening: "Let's explore this new topic..."
Problem: Doesn't reference Lesson N-1
→ Rewrite opening only; don't cut words

Example 3: Both

Word count: 1950 (23% over)
Opening: Generic, missing prior lesson reference
→ Phase 1: Rewrite opening (identify anchor from Lesson N-1)
→ Phase 2: Cut words to 23% (context-aware)

询问用户或查看Gate 4的失败提示：

失败类型	问题	操作
字数问题	"课程是否超过目标字数（通常为1500词）？"	计算精确的删减比例
连贯性问题	"开篇是否提及上一课程？"	仅重写开篇
两者皆有	"字数和连贯性都有问题？"	分两阶段处理

诊断示例:

示例1：仅字数问题

Content: 1950 words, Target: 1500
Excess: 450 words
% to cut: (450 / 1950) × 100 = 23%
→ CUT EXACTLY 23%, not generic 15-20%

示例2：仅连贯性问题

Opening: "Let's explore this new topic..."
Problem: Doesn't reference Lesson N-1
→ Rewrite opening only; don't cut words

示例3：两者皆有

Word count: 1950 (23% over)
Opening: Generic, missing prior lesson reference
→ Phase 1: Rewrite opening (identify anchor from Lesson N-1)
→ Phase 2: Cut words to 23% (context-aware)

Step 1: Assess Content Layer (Context-Aware Cutting)

步骤1：评估内容层级（基于上下文的删减）

Read the lesson's frontmatter to determine layer:

Layer	Cutting Strategy
L1 (Manual)	Keep foundational explanations; cut elaboration
L2 (AI-Collaboration)	Keep Try With AI sections (core); cut narrative padding
L3 (Intelligence)	Keep pattern insights; cut explanatory scaffolding
L4 (Spec-Driven)	Keep specification details; cut conceptual scaffolding

查看课程的frontmatter以确定层级：

层级	删减策略
L1（手动）	保留基础解释；删减冗余阐述
L2（AI协作）	保留Try With AI板块（核心内容）；删减叙事性铺垫
L3（进阶）	保留模式洞察；删减解释性框架内容
L4（规范驱动）	保留规范细节；删减概念性框架内容

The Refinement Procedure (Layer-Aware)

优化流程（基于层级）

Phase 1: The Connection Builder (Continuity Fix)

阶段1：关联性构建（修复连贯性）

Do this FIRST if opening is generic.

Formula:

markdown

In [Previous Lesson], you [SPECIFIC OUTCOME from Lesson N-1].
Now, we will [CONNECT outcome to new goal] by [STRATEGY].

Validation:

Opening references Lesson N-1 by name
Specific outcome (not generic "learned about...")
Clear connection shows why this lesson matters (builds on N-1)

After fixing: Proceed to Fluff Cutter if word count also fails.

如果开篇表述泛化，请先执行此步骤。

公式:

markdown

在[上一课程]中，你完成了[课程N-1的具体成果]。
现在，我们将[把该成果与新目标关联]，通过[具体策略]实现。

验证项:

开篇提及课程N-1的名称
包含具体成果（而非泛泛的"学习了..."）
清晰说明本课程的重要性（基于课程N-1的延伸）

修复后：如果同时存在字数问题，继续执行冗余内容精简步骤。

Phase 2: The Fluff Cutter (Word Count Fix)

阶段2：冗余内容精简（修复字数问题）

Apply layer-specific cuts in this order:

FOR ALL LAYERS:

Delete redundant "Why This Matters" sections
- Keep ONLY if it reveals non-obvious insight
- If same point made in text AND in "Why This Matters" → delete WTM
Merge repeated examples
- Find duplicate explanations
- Keep first, delete second
Tighten transitions between sections
- Replace "As we discussed earlier, X..." with direct reference

FOR L1-L2 ONLY (students still building foundation): 4. Reduce "Try With AI" sections to exactly 2 prompts

Keep foundational + one advanced
Delete exploratory extras

Keep educational scaffolding (explanations, examples)

FOR L3-L4 ONLY (students ready for advanced patterns): 4. Trim narrative scaffolding

Keep pattern insights and rules
Delete "why this matters philosophically"

Remove beginner-level explanations
- Assume students understand fundamentals

FOR ALL LAYERS: 6. One Analogy Rule: Keep the BEST analogy for the concept; delete redundant ones 7. Merge Tables/Text: Use ONE format (table OR prose), never both 8. Reduce Examples: Keep 2-3 best; delete "also consider..." 9. Tighten Lists: Convert 5-item lists to 3 core items

Verification:

Word count after cuts: [TARGET ± 5%]
No L1 content cut from L1 lessons
No pattern insights lost from L3-L4 lessons
Try With AI: 2 prompts if L1-L2, keep all if L3-L4

按以下顺序应用基于层级的删减策略:

所有层级通用:

删除冗余的"Why This Matters"板块
- 仅保留能揭示非显而易见见解的内容
- 如果正文和"Why This Matters"板块表述相同，删除该板块
合并重复示例
- 找出重复的解释内容
- 保留第一个，删除第二个
精简板块间的过渡语句
- 用直接指代替换"正如我们之前讨论的，X..."这类表述

仅适用于L1-L2层级（学员仍在构建基础）: 4. 将"Try With AI"板块精简为恰好2个提示词

保留基础提示词+1个进阶提示词
删除探索性的额外提示词

保留教学框架内容（解释、示例）

仅适用于L3-L4层级（学员已准备好学习进阶模式）: 4. 精简叙事性框架内容

保留模式洞察和规则
删除"从哲学角度看为何重要"这类内容

移除入门级解释内容
- 假设学员已掌握基础知识

所有层级通用: 6. 单一类比规则：保留针对概念的最佳类比；删除冗余类比 7. 合并表格/文本：仅使用一种格式（表格或 prose），切勿同时使用 8. 精简示例：保留2-3个最佳示例；删除"也可以考虑..."这类内容 9. 精简列表：将5项列表压缩为3项核心内容

验证项:

删减后的字数：[目标值±5%]
未删减L1层级的核心内容
未删减L3-L4层级的模式洞察内容
Try With AI：L1-L2层级保留2个提示词，L3-L4层级保留全部

Phase 3: Post-Refinement Validation (CRITICAL)

阶段3：优化后验证（关键步骤）

After applying fixes, verify the content now PASSES Gate 4:

✓ Word Count Check:
  Current: [X] words
  Target: [target_from_spec]
  Status: [PASS if ≤target ± 5%, FAIL if over]

✓ Continuity Check:
  Opening references Lesson [N-1]? [YES/NO]
  Specific outcome mentioned? [YES/NO]
  Connection to new lesson clear? [YES/NO]

✓ Layer Appropriateness:
  No foundational cuts from L1-L2? [YES/NO]
  No pattern insight loss from L3-L4? [YES/NO]

✓ Content Integrity:
  Removed examples still explained elsewhere? [YES/NO]
  Cut sections non-essential? [YES/NO]

NEXT STEP RECOMMENDATION:

"Refined content is ready.

Word count: [after] (target: ≤[target])
Continuity: Now references Lesson [N-1]

Recommend re-submitting to acceptance-auditor for Gate 4 re-validation.
Command: [provide re-validation instruction]"

应用修复后，验证内容是否已通过Gate 4:

✓ 字数检查:
  当前字数: [X]词
  目标字数: [规范中的目标值]
  状态: [如果≤目标值±5%则通过，否则不通过]

✓ 连贯性检查:
  开篇提及课程[N-1]？[是/否]
  提及具体成果？[是/否]
  与新课程的关联是否清晰？[是/否]

✓ 层级适配性:
  未删减L1-L2层级的基础内容？[是/否]
  未删减L3-L4层级的模式洞察内容？[是/否]

✓ 内容完整性:
  被删除的示例是否在其他地方有解释？[是/否]
  被删减的板块是否非必要？[是/否]

下一步建议:

"优化后的内容已准备就绪。

当前字数: [优化后]
连贯性: 已提及课程[N-1]

建议重新提交至acceptance-auditor进行Gate 4重新验证。
指令: [提供重新验证的具体说明]"

Output Format

输出格式

markdown

undefined

markdown

undefined

Refinement Report: [Lesson Name]

优化报告: [课程名称]

Diagnosis

诊断结果

Issue Found: [Word count | Continuity | Both] Layer: [L1/L2/L3/L4]

发现的问题: [字数问题 | 连贯性问题 | 两者皆有] 层级: [L1/L2/L3/L4]

Metrics

指标

Metric	Before	After	Target	Status
Word Count	1950	1485	≤1500	✅ PASS
Continuity	Generic opening	References Lesson 2	Specific reference	✅ PASS

指标	优化前	优化后	目标值	状态
字数	1950	1485	≤1500	✅ 通过
连贯性	泛化开篇	提及课程2	具体关联	✅ 通过

Fixes Applied

应用的修复措施

Phase 1: Rewrote opening to reference "booking-agent implementation" from Lesson 2
Phase 2: Deleted 240 words using layer-aware cuts:
- Removed redundant "Why This Matters" section (line 45, 120 words)
- Merged duplicate example (lines 67-89, 85 words)
- Cut 1 extra "Try With AI" prompt (35 words)
Phase 3: Validated word count and continuity

阶段1: 重写开篇，提及课程2中的"booking-agent实现"内容
阶段2: 基于层级删减了240词:
- 删除了冗余的"Why This Matters"板块（第45行，120词）
- 合并了重复示例（第67-89行，85词）
- 删减了1个额外的"Try With AI"提示词（35词）
阶段3: 验证了字数和连贯性

Ready for Re-validation

已准备好重新验证

✅ Word count: 1485 (≤1500) ✅ Continuity: Opening references Lesson 2 ✅ Layer integrity: All L2 AI examples preserved

Next: Re-submit to acceptance-auditor for Gate 4 validation

✅ 字数: 1485（≤1500） ✅ 连贯性: 开篇提及课程2 ✅ 层级完整性: 所有L2层级的AI示例均被保留

下一步: 重新提交至acceptance-auditor进行Gate 4验证

Refined Content

优化后的内容

[Full refined lesson content]

undefined

[完整的优化后课程内容]

undefined