fix: 修复 anthropic 煞笔的四个 bug (#352)

* fix: 移除文件编辑前必须先读取的限制

移除 FileEditTool 和 FileWriteTool 中的 "read before edit" 校验，
允许直接编辑未读取过的文件。保留文件修改过期检测。

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

* docs: 更新 teach-me 自动写 note 笔记的功能

* fix: 修复 DeepSeek V4 reasoning_content 回传导致的 400 错误

- 扩大模型名称检测范围，匹配所有 deepseek 模型（V4、R1 等）
- 始终保留 thinking blocks 为 reasoning_content 回传给 API
- 移除有 bug 的 turn boundary 剥离逻辑

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

* fix: Opus 4.6/4.7 默认推理 effort 从 medium 改为 high

Pro 和 Max/Team 订阅者的 Opus 默认 effort 之前被降级为 medium，
导致用户感知模型「变笨」。恢复为 high。

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

* fix: 移除 thinkingClearLatched sticky-on 机制

空闲超过 1 小时后 thinkingClearLatched 会被触发且永不重置，
导致每轮 API 调用都清除 thinking 历史。完整移除该 latch 机制，
clearAllThinking 硬编码为 false。

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

* fix: 移除 numeric_length_anchors 系统指令

删除「工具调用间文字 ≤25 词、最终回复 ≤100 词」的硬性限制。
ablation 测试显示该约束使整体智能下降 3%。

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

* fix: 修复测试中 reasoning_content 类型断言

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>

src/services/api/claude.ts

@@ -124,14 +124,12 @@ import {
   getPromptCache1hAllowlist,
   getPromptCache1hEligible,
   getSessionId,
-  getThinkingClearLatched,
   setAfkModeHeaderLatched,
   setCacheEditingHeaderLatched,
   setFastModeHeaderLatched,
   setLastMainRequestId,
   setPromptCache1hAllowlist,
   setPromptCache1hEligible,
-  setThinkingClearLatched,
 } from 'src/bootstrap/state.js'
 import {
   AFK_MODE_BETA_HEADER,
@@ -1492,20 +1490,6 @@ async function* queryModel(
     }
   }
 
-  // Only latch from agentic queries so a classifier call doesn't flip the
-  // main thread's context_management mid-turn.
-  let thinkingClearLatched = getThinkingClearLatched() === true
-  if (!thinkingClearLatched && isAgenticQuery) {
-    const lastCompletion = getLastApiCompletionTimestamp()
-    if (
-      lastCompletion !== null &&
-      Date.now() - lastCompletion > CACHE_TTL_1HOUR_MS
-    ) {
-      thinkingClearLatched = true
-      setThinkingClearLatched(true)
-    }
-  }
-
   const effort = resolveAppliedEffort(options.model, options.effortValue)
 
   if (feature('PROMPT_CACHE_BREAK_DETECTION')) {
@@ -1684,7 +1668,7 @@ async function* queryModel(
     const contextManagement = getAPIContextManagement({
       hasThinking,
       isRedactThinkingActive: betasParams.includes(REDACT_THINKING_BETA_HEADER),
-      clearAllThinking: thinkingClearLatched,
+      clearAllThinking: false,
     })
 
     const enablePromptCaching =