Menu
The review-first prompt is selected by default. You can switch to a direct command or download a local copy.
Read SKILL.md and any companion files shown by SkillsMP before deciding whether to install.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
A direct command skips the review prompt. Inspect the source before running it.
npx skills add https://github.com/loulanyue/awesome-claude-notes --skill continuous-learningThe command stays on one line. Scroll horizontally to inspect it before copying.
Prefer a local copy? Download the files currently available to SkillsMP.
Showing SKILL.md
Head-to-head comparison of coding agents (Claude Code, Aider, Codex, etc.) on custom tasks with pass rate, cost, time, and consistency metrics
Design and optimize AI agent action spaces, tool definitions, and observation formatting for higher completion rates.
Operate as an agentic engineer using eval-first execution, decomposition, and cost-aware model routing.
Based on SOC occupation classification
| name | continuous-learning |
| description | 自动从Claude Code会话中提取可重复使用的模式，并将其保存为学习到的技能以供将来使用。 |
| source_path | skills/continuous-learning/SKILL.md |
| origin | ECC |
自动评估 Claude Code 会话的结尾，以提取可重用的模式，这些模式可以保存为学习到的技能。
~/.claude/skills/learned/ 中审查或整理已学习的技能此技能作为 停止钩子 在每个会话结束时运行：
~/.claude/skills/learned/编辑 config.json 以进行自定义：
{
"min_session_length": 10,
"extraction_threshold": "medium",
"auto_approve": false,
"learned_skills_path": "~/.claude/skills/learned/",
"patterns_to_detect": [
"error_resolution",
"user_corrections",
"workarounds",
"debugging_techniques",
"project_specific"
],
"ignore_patterns": [
"simple_typos",
"one_time_fixes",
"external_api_issues"
]
}
| 模式 | 描述 |
|---|---|
error_resolution | 特定错误是如何解决的 |
user_corrections | 来自用户纠正的模式 |
workarounds | 框架/库特殊性的解决方案 |
debugging_techniques | 有效的调试方法 |
project_specific | 项目特定的约定 |
添加到你的 ~/.claude/settings.json 中：
{
"hooks": {
"Stop": [{
"matcher": "*",
"hooks": [{
"type": "command",
"command": "~/.claude/skills/continuous-learning/evaluate-session.sh"
}]
}]
}
}
/learn 命令 - 在会话中手动提取模式Homunculus v2 采用了更复杂的方法：
| 功能 | 我们的方法 | Homunculus v2 |
|---|---|---|
| 观察 | 停止钩子（会话结束时） | PreToolUse/PostToolUse 钩子（100% 可靠） |
| 分析 | 主上下文 | 后台代理 (Haiku) |
| 粒度 | 完整技能 | 原子化的“本能” |
| 置信度 | 无 | 0.3-0.9 加权 |
| 演进 | 直接到技能 | 本能 → 集群 → 技能/命令/代理 |
| 共享 | 无 | 导出/导入本能 |
来自 homunculus 的关键见解：
"v1 依赖技能来观察。技能是概率性的——它们触发的概率约为 50-80%。v2 使用钩子进行观察（100% 可靠），并以本能作为学习行为的原子单元。"
参见：docs/continuous-learning-v2-spec.md 以获取完整规范。