Continuous Learning

affaan-m/ECC/docs/zh-TW/skills/continuous-learning

by affaan-mef648e01899ba3e8dc6371642deaaf64b4477775No license275K starsListed Oct 9, 2026Updated Oct 9, 2026Repository updated 4 days ago

Automatically extract reusable patterns from Claude Code sessions and save them as learned skills for future use.

Instructions onlyAI & Agents

AI-generated overview

Extracts reusable patterns from Claude Code sessions and saves them as learned skills for later use.

What it does: This skill runs as a Stop hook at the end of a Claude Code session. It evaluates whether the session has enough messages, detects extractable patterns such as error resolutions, user corrections, workarounds, debugging techniques and project-specific conventions, and saves useful patterns as learned skills under a configured directory. It is configuration-driven through a config file and hook settings, and it also documents a comparison with an alternative approach and possible future enhancements.
When to use it: Use it when you want Claude Code sessions to automatically accumulate reusable knowledge without manual effort. It suits users who want recurring fixes, corrections and debugging approaches captured as skills at session end. It is also relevant if you are evaluating hook-based learning designs or planning a more granular version.
Requirements: Requires Claude Code with hook support and a Stop hook entry in the settings file, plus a config file for thresholds, paths and pattern lists. It writes learned skills to a local skills directory. No scripts are included; the skill is instructions only.

持續學習技能

自動評估 Claude Code 工作階段結束時的內容，提取可重用模式並儲存為學習技能。

運作方式

此技能作為 Stop hook 在每個工作階段結束時執行：

工作階段評估：檢查工作階段是否有足夠訊息（預設：10+ 則）
模式偵測：從工作階段識別可提取的模式
技能提取：將有用模式儲存到 ~/.claude/skills/learned/

設定

編輯 config.json 以自訂：

json

{  "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"  ]}

{  "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"  ]}

模式類型

模式	描述
`error_resolution`	特定錯誤如何被解決
`user_corrections`	來自使用者修正的模式
`workarounds`	框架/函式庫怪異問題的解決方案
`debugging_techniques`	有效的除錯方法
`project_specific`	專案特定慣例

Hook 設定

新增到你的 ~/.claude/settings.json：

json

{  "hooks": {    "Stop": [{      "matcher": "*",      "hooks": [{        "type": "command",        "command": "~/.claude/skills/continuous-learning/evaluate-session.sh"      }]    }]  }}

{  "hooks": {    "Stop": [{      "matcher": "*",      "hooks": [{        "type": "command",        "command": "~/.claude/skills/continuous-learning/evaluate-session.sh"      }]    }]  }}

為什麼用 Stop Hook？

輕量：工作階段結束時只執行一次
非阻塞：不會為每則訊息增加延遲
完整上下文：可存取完整工作階段記錄

比較筆記（研究：2025 年 1 月）

vs Homunculus

Homunculus v2 採用更複雜的方法：

功能	我們的方法	Homunculus v2
觀察	Stop hook（工作階段結束）	PreToolUse/PostToolUse hooks（100% 可靠）
分析	主要上下文	背景 agent（Haiku）
粒度	完整技能	原子「本能」
信心	無	0.3-0.9 加權
演化	直接到技能	本能 → 聚類 → 技能/指令/agent
分享	無	匯出/匯入本能

來自 homunculus 的關鍵見解：

"v1 依賴技能進行觀察。技能是機率性的——它們觸發約 50-80% 的時間。v2 使用 hooks 進行觀察（100% 可靠），並以本能作為學習行為的原子單位。"

潛在 v2 增強

基於本能的學習 - 較小的原子行為，帶信心評分
背景觀察者 - Haiku agent 並行分析
信心衰減 - 如果被矛盾則本能失去信心
領域標記 - code-style、testing、git、debugging 等
演化路徑 - 將相關本能聚類為技能/指令

參見：docs/continuous-learning-v2-spec.md 完整規格。

Source and attribution

Source:affaan-m/ECCindocs/zh-TW/skills/continuous-learningat commitef648e0

License: No license

Content belongs to its original authors. SourceWeft indexes it from a public repository.

Report or request removal

Continuous Learning

持續學習技能

運作方式

設定

模式類型

Hook 設定

為什麼用 Stop Hook？

相關

比較筆記（研究：2025 年 1 月）

vs Homunculus

潛在 v2 增強

Source and attribution

More from affaan-m/ECC

Unified Notifications Ops

Verification Loop

Ui To Vue

Unified Memory

Visa Doc Translate

Team Builder

More in AI & Agents

Ppt Template Creator

Skill Development

Plugin Structure

Command Development

Claude Md Improver

Find Skills