工作原理

1. 自动记录（通过钩子）

PostToolUse → enhanced-failure-logger.js → 记录带上下文的失败
UserMessage → correction-detector.js → 检测 "错误/实际上/再试一次"
Response → capability-tracker.js → 检测未满足的请求

2. 学习聚合

所有学习成果流向 logs/learnings.jsonl：

{
  "timestamp": "2026-01-26T12:00:00Z",
  "type": "user_correction|tool_failure|missing_capability",
  "category": "factual_error|command_failed|web_browsing",
  "description": "学到的内容",
  "source": "哪个检测器"
}

3. 会话开始回顾

每次会话开始时，显示最近的学习成果：

=== 学习回顾 ===
[学习成果] 总计 47 条
[最近]
  • [tool_failure] Bash 失败：超时 - WebFetch 到外部 API...
  • [user_correction] 用户纠正："不，使用另一个文件..."
[纠正] 记录了 12 条用户纠正
[能力差距] 最常请求的：
  • 发送电子邮件 (5次)
  • 浏览网页 (3次)

4. QAVR 集成

成功的学习成果会提升相关记忆的 Q 值，改善未来的检索。

手动命令

查看学习成果

# 显示所有学习成果
cat ~/.claude/logs/learnings.jsonl | tail -20

# 仅显示纠正
cat ~/.claude/logs/corrections.jsonl | jq -s 'group_by(.correction_type) | map({type: .[0].correction_type, count: length})'

# 显示能力差距报告
node ~/.claude/scripts/hooks/capability-tracker.js --report

测试检测

# 测试纠正检测器
node ~/.claude/scripts/hooks/correction-detector.js

# 测试失败记录器
node ~/.claude/scripts/hooks/enhanced-failure-logger.js

# 测试能力追踪器
node ~/.claude/scripts/hooks/capability-tracker.js

学习类别

纠正类型

• factual_error - 提供了错误信息
• retry_request - 用户要求重试
• misunderstanding - 误解了请求
• failed_solution - 解决方案无效

失败类型

• permission_error - 访问被拒绝
• not_found - 文件/资源缺失
• timeout - 操作超时
• network_error - 连接问题
• syntax_error - 无效语法
• api_error - 外部 API 失败
• command_failed - Shell 命令失败
• agent_failed - 子代理失败

能力类别

• web_browsing - 互联网访问请求
• image_processing - 图像/照片处理
• communication - 电子邮件/消息
• database_access - SQL/数据库查询
• external_api - 第三方服务
• memory_persistence - 长期记忆

配置

在 settings.json 中，这些钩子启用自我改进：

{
  "hooks": {
    "SessionStart": [...],  // 回顾学习成果
    "PostToolUse": [
      {"matcher": "Bash", "hooks": [{"command": "enhanced-failure-logger.js"}]},
      {"matcher": "Task", "hooks": [{"command": "enhanced-failure-logger.js"}]}
    ]
  }
}

优势

1. 从错误中学习 - 避免重复相同的错误
2. 理解用户偏好 - 追踪纠正的含义
3. 识别技能差距 - 知道需要构建哪些功能
4. 随时间改进 - QAVR 排名随反馈提升
5. 上下文持久性 - 学习成果在会话重启后保留

与其他技能的集成

文件

~/.claude/
├── logs/
│   ├── learnings.jsonl           # 所有学习成果
│   ├── corrections.jsonl         # 用户纠正
│   ├── failures_detailed.jsonl   # 增强的失败日志
│   ├── missing_capabilities.jsonl # 能力请求
│   └── capability_gaps.json      # 聚合的差距
└── scripts/hooks/
    ├── correction-detector.js
    ├── enhanced-failure-logger.js
    ├── capability-tracker.js
    └── session-start.js (回顾学习成果)

每周安装次数

121

仓库

kimasplund/claw…lls-pack

首次出现

2026年2月9日

安全审计

Gen Agent Trust HubFailSocketPassSnykWarn

安装于

gemini-cli117

github-copilot116

codex116

opencode116

cursor116

kimi-cli115

How It Works

1. Automatic Logging (via hooks)

PostToolUse → enhanced-failure-logger.js → logs failures with context
UserMessage → correction-detector.js → detects "wrong/actually/try again"
Response → capability-tracker.js → detects unfulfilled requests

2. Learning Aggregation

All learnings flow to logs/learnings.jsonl:

{
  "timestamp": "2026-01-26T12:00:00Z",
  "type": "user_correction|tool_failure|missing_capability",
  "category": "factual_error|command_failed|web_browsing",
  "description": "what was learned",
  "source": "which detector"
}

3. Session Start Review

On each session start, recent learnings are shown:

=== Learning Review ===
[Learnings] 47 total entries
[Recent]
  • [tool_failure] Bash failed: timeout - WebFetch to external API...
  • [user_correction] User corrected: "No, use the other file..."
[Corrections] 12 user corrections logged
[Capability Gaps] Top requested:
  • send emails (5x)
  • browse web (3x)

4. QAVR Integration

Successful learnings boost Q-values for related memories, improving future retrieval.

Manual Commands

Review Learnings

# Show all learnings
cat ~/.claude/logs/learnings.jsonl | tail -20

# Show corrections only
cat ~/.claude/logs/corrections.jsonl | jq -s 'group_by(.correction_type) | map({type: .[0].correction_type, count: length})'

# Show capability gaps report
node ~/.claude/scripts/hooks/capability-tracker.js --report

Test Detection

# Test correction detector
node ~/.claude/scripts/hooks/correction-detector.js

# Test failure logger
node ~/.claude/scripts/hooks/enhanced-failure-logger.js

# Test capability tracker
node ~/.claude/scripts/hooks/capability-tracker.js

Learning Categories

Correction Types

• factual_error - Wrong information provided
• retry_request - User asked to try again
• misunderstanding - Misinterpreted the request
• failed_solution - Solution didn't work

Failure Types

• permission_error - Access denied
• not_found - File/resource missing
• timeout - Operation timed out
• network_error - Connection issues
• syntax_error - Invalid syntax
• api_error - External API failed
• command_failed - Shell command failed
• agent_failed - Subagent failed

Capability Categories

• web_browsing - Internet access requests
• image_processing - Image/photo handling
• communication - Email/messaging
• database_access - SQL/database queries
• external_api - Third-party services
• memory_persistence - Long-term memory

Configuration

In settings.json, these hooks enable self-improvement:

{
  "hooks": {
    "SessionStart": [...],  // Reviews learnings
    "PostToolUse": [
      {"matcher": "Bash", "hooks": [{"command": "enhanced-failure-logger.js"}]},
      {"matcher": "Task", "hooks": [{"command": "enhanced-failure-logger.js"}]}
    ]
  }
}

Benefits

1. Learn from mistakes - Don't repeat the same errors
2. Understand user preferences - Track what corrections mean
3. Identify skill gaps - Know what features to build
4. Improve over time - QAVR ranking gets better with feedback
5. Context persistence - Learnings survive session restarts

Integration with Other Skills

Files

~/.claude/
├── logs/
│   ├── learnings.jsonl           # All learnings
│   ├── corrections.jsonl         # User corrections
│   ├── failures_detailed.jsonl   # Enhanced failure logs
│   ├── missing_capabilities.jsonl # Capability requests
│   └── capability_gaps.json      # Aggregated gaps
└── scripts/hooks/
    ├── correction-detector.js
    ├── enhanced-failure-logger.js
    ├── capability-tracker.js
    └── session-start.js (reviews learnings)

Weekly Installs

121

Repository

kimasplund/claw…lls-pack

First Seen

Feb 9, 2026

Security Audits

Gen Agent Trust HubFailSocketPassSnykWarn

Installed on

gemini-cli117

github-copilot116

codex116

opencode116

cursor116

kimi-cli115