fix: 5 处准确性和适用性问题修正

1. cheatsheet 删除 "Esc Esc 回滚 checkpoint" 说明
   这个描述在不同 CLI 版本行为不一，无法确定准确含义。
   保留 "Esc 打断当前生成" 这条稳定的说明。

2. Aider 示例模型名从 3.5 系列升到 4.5 系列
   3.5 是旧版本。统一项目里 13 处引用到当前 4.x 模型。
   涉及: cheatsheet / pitfalls/aider / aider/README / templates

3. Copilot 陷阱 7 措辞修正
   "静默降级" → "排队或拒绝"，更准确反映 Free 版到量后的实际行为。

4. 英文 README 新手推荐从 Trae 改为 Copilot/Cursor
   Trae 是字节产品，中文界面、国内直连是优势，对国际英文读者
   反而是门槛。中文 README 保留 Trae 推荐不变。

5. .lychee.toml 移除两个不确定的 key
   github_token = "" 和 include_fragments 可能因 lychee 版本
   差异报错。移除后依赖默认行为，更稳妥。

Author: jnMetaCode

.lychee.toml

@@ -46,12 +46,6 @@ exclude = [
   "^https?://qm\\.qq\\.com",
 ]
 
-# 对私有仓库的 GitHub URL 跳过认证检查（公开仓库不影响）
-github_token = ""
-
-# 跳过 fragment（#section）校验，避免跨仓 anchor 误报
-include_fragments = false
-
 # 缓存
 cache = true
 max_cache_age = "1d"

README.en.md

@@ -27,7 +27,7 @@
 
 | You are... | Recommended reading order |
 |------------|---------------------------|
-| Total newcomer to AI coding | [Cheatsheet](cheatsheet.en.md) → [Trae](trae/) (free) → [Prompt Engineering](common/prompting.en.md) |
+| Total newcomer to AI coding | [Cheatsheet](cheatsheet.en.md) → [Copilot](copilot/) (free tier) or [Cursor](cursor/) (free tier) → [Prompt Engineering](common/prompting.en.md) |
 | Frontend / daily coding | [Cursor](cursor/) → [Prompt Engineering](common/prompting.en.md) → [Scenarios](workflows/scenarios.en.md) |
 | Backend / refactoring / big projects | [Claude Code](claude-code/) → [Task Decomposition](common/task-decomposition.en.md) → [Scenarios](workflows/scenarios.en.md) |
 | Migrating from Copilot | [Copilot](copilot/) → [Claude Code](claude-code/) (compare) → [Tool Selection](workflows/tool-selection.en.md) |

aider/README.en.md

@@ -38,7 +38,7 @@ cd /your/project
 aider
 
 # Specify a model
-aider --model claude-3-5-sonnet
+aider --model claude-sonnet-4-5
 
 # Use DeepSeek (cheaper)
 aider --model deepseek/deepseek-chat
@@ -116,7 +116,7 @@ aider --commit-prefix "[ai] "
 
 ```bash
 # Complex architecture design — use the strongest model
-aider --model claude-3-5-sonnet
+aider --model claude-sonnet-4-5
 /architect Design a microservices split plan
 
 # Daily coding — use a cost-effective model
@@ -136,7 +136,7 @@ aider --model ollama/qwen2.5-coder
 
 ```yaml
 # .aider.conf.yml
-model: claude-3-5-sonnet
+model: claude-sonnet-4-5
 auto-commits: true
 auto-lint: true
 auto-test: true

aider/README.md

@@ -36,7 +36,7 @@ cd /your/project
 aider
 
 # 指定模型
-aider --model claude-3-5-sonnet
+aider --model claude-sonnet-4-5
 
 # 用 DeepSeek（便宜）
 aider --model deepseek/deepseek-chat
@@ -114,7 +114,7 @@ aider --commit-prefix "[ai] "
 
 ```bash
 # 复杂架构设计 — 用最强模型
-aider --model claude-3-5-sonnet
+aider --model claude-sonnet-4-5
 /architect 设计微服务拆分方案
 
 # 日常编码 — 用性价比模型
@@ -134,7 +134,7 @@ aider --model ollama/qwen2.5-coder
 
 ```yaml
 # .aider.conf.yml
-model: claude-3-5-sonnet
+model: claude-sonnet-4-5
 auto-commits: true
 auto-lint: true
 auto-test: true

aider/templates/aider.conf.yml

@@ -2,7 +2,7 @@
 # 复制到项目根目录，命名为 .aider.conf.yml
 
 # 模型配置
-model: claude-3-5-sonnet
+model: claude-sonnet-4-5
 # 其他选择：
 # model: deepseek/deepseek-chat      # 便宜，适合日常编码
 # model: ollama/qwen2.5-coder        # 免费，需要本地跑 Ollama

cheatsheet.en.md

@@ -77,7 +77,7 @@ claude -p "task" --output-format json   # Headless
 # Inside the session
 /compact                        # Compress context
 /plan                           # Enter plan mode
-Esc Esc                         # Roll back to checkpoint
+Esc                             # Interrupt current generation
 ```
 
 ### Cursor
@@ -141,7 +141,7 @@ Steering load modes:
 ```bash
 # Install & run
 pip install aider-chat
-aider --model claude-3-5-sonnet
+aider --model claude-sonnet-4-5
 aider --model deepseek/deepseek-chat   # Cheap
 aider --model ollama/qwen2.5-coder     # Free, local
 

cheatsheet.md

@@ -75,7 +75,7 @@ claude -p "任务" --output-format json   # headless 模式
 # 交互中
 /compact                        # 压缩上下文
 /plan                           # 进入 plan 模式
-Esc Esc                         # 回滚到 checkpoint
+Esc                             # 打断当前生成
 ```
 
 ### Cursor
@@ -139,7 +139,7 @@ Steering 加载模式：
 ```bash
 # 安装与启动
 pip install aider-chat
-aider --model claude-3-5-sonnet
+aider --model claude-sonnet-4-5
 aider --model deepseek/deepseek-chat   # 便宜
 aider --model ollama/qwen2.5-coder     # 本地免费
 

pitfalls/aider.en.md

@@ -85,9 +85,9 @@ LLMs vary wildly in instruction-following. Same `/architect design a WebSocket n
 Tier in `.aider.conf.yml`:
 
 ```yaml
-model: claude-3-5-sonnet           # Primary
+model: claude-sonnet-4-5           # Primary
 weak-model: deepseek/deepseek-chat # Weak tasks (commit messages, simple completion)
-editor-model: claude-3-5-haiku     # Editor-mode completion
+editor-model: claude-haiku-4-5     # Editor-mode completion
 ```
 
 Or switch **by task type**:

pitfalls/aider.md

@@ -83,9 +83,9 @@ Aider 只把 `/add` 过的文件真正**读进上下文**。Map 模式会索引
 配置分层（`.aider.conf.yml`）：
 
 ```yaml
-model: claude-3-5-sonnet           # 主模型
+model: claude-sonnet-4-5           # 主模型
 weak-model: deepseek/deepseek-chat # 弱任务（commit message、简单补全）
-editor-model: claude-3-5-haiku     # 编辑器补全
+editor-model: claude-haiku-4-5     # 编辑器补全
 ```
 
 或**按任务类型**切：

pitfalls/copilot.en.md

@@ -185,11 +185,11 @@ Either standardize IDE team-wide, or accept JetBrains being a step behind.
 
 **Symptom**
 - Copilot Free worked fine; one day completion latency spikes
-- Chat gets slow, "thinking..." for tens of seconds
-- No error, just slow, with dropped quality (like a downgraded model)
+- Chat requests queue or get outright rejected ("usage limit reached"-style messages)
+- Completions slow down, fewer suggestions
 
 **Cause**
-Copilot Free has monthly caps and rate limits. Hitting them triggers **silent degradation**, not an explicit error. Enterprise users may hit org-level quotas similarly.
+Copilot Free has monthly completion and Chat quotas. As you approach them you get rate-limited; once exceeded, requests are rejected. The notification in VS Code isn't always prominent, so it's easy to miss. Enterprise users may hit org-level quotas similarly.
 
 **Recovery**
 - GitHub Account → Copilot settings to check usage