docs: 完成所有文档的中文翻译并应用到项目

2026-03-22 06:20:10 +00:00 · 2026-01-28 00:12:54 +08:00
parent 0ced59a26b
commit e133f58e1c
76 changed files with 6808 additions and 6170 deletions
--- a/skills/continuous-learning-v2/SKILL.md
+++ b/skills/continuous-learning-v2/SKILL.md
@@ -1,27 +1,27 @@
 ---
 name: continuous-learning-v2
-description: Instinct-based learning system that observes sessions via hooks, creates atomic instincts with confidence scoring, and evolves them into skills/commands/agents.
+description: 基于直觉的学习系统，通过钩子（hooks）观察会话，创建带有置信度评分的原子直觉（atomic instincts），并将其演化为技能/命令/智能体。
 version: 2.0.0
 ---

-# Continuous Learning v2 - Instinct-Based Architecture
+# 持续学习 v2 - 基于直觉的架构 (Instinct-Based Architecture)

-An advanced learning system that turns your Claude Code sessions into reusable knowledge through atomic "instincts" - small learned behaviors with confidence scoring.
+一个高级学习系统，通过原子“直觉（instincts）”——带有置信度评分的小型习得行为，将你的 Claude Code 会话转化为可复用的知识。

-## What's New in v2
+## v2 版本新特性

-| Feature | v1 | v2 |
+| 特性 | v1 | v2 |
 |---------|----|----|
-| Observation | Stop hook (session end) | PreToolUse/PostToolUse (100% reliable) |
-| Analysis | Main context | Background agent (Haiku) |
-| Granularity | Full skills | Atomic "instincts" |
-| Confidence | None | 0.3-0.9 weighted |
-| Evolution | Direct to skill | Instincts → cluster → skill/command/agent |
-| Sharing | None | Export/import instincts |
+| 观察 (Observation) | Stop 钩子（会话结束） | PreToolUse/PostToolUse (100% 可靠) |
+| 分析 (Analysis) | 主上下文 (Main context) | 后台智能体 (Haiku) |
+| 粒度 (Granularity) | 完整技能 (Full skills) | 原子“直觉（instincts）” |
+| 置信度 (Confidence) | 无 | 0.3-0.9 加权 |
+| 演化 (Evolution) | 直接转换为技能 | 直觉 → 聚类 → 技能/命令/智能体 |
+| 共享 (Sharing) | 无 | 导出/导入直觉 |

-## The Instinct Model
+## 直觉模型 (The Instinct Model)

-An instinct is a small learned behavior:
+“直觉（instinct）”是一个小型的习得行为：

 ```yaml
 ---
@@ -32,44 +32,44 @@ domain: "code-style"
 source: "session-observation"
 ---

-# Prefer Functional Style
+# 偏好函数式风格 (Prefer Functional Style)

-## Action
-Use functional patterns over classes when appropriate.
+## 操作 (Action)
+在合适的时候，优先使用函数式模式而非类（class）。

-## Evidence
- Observed 5 instances of functional pattern preference
- User corrected class-based approach to functional on 2025-01-15
+## 证据 (Evidence)
+- 观察到 5 次偏好函数式模式的实例
+- 用户在 2025-01-15 将基于类的方法修正为函数式
 ```

-**Properties:**
- **Atomic** — one trigger, one action
- **Confidence-weighted** — 0.3 = tentative, 0.9 = near certain
- **Domain-tagged** — code-style, testing, git, debugging, workflow, etc.
- **Evidence-backed** — tracks what observations created it
+**属性：**
+- **原子性 (Atomic)** — 一个触发器，一个动作
+- **置信度加权 (Confidence-weighted)** — 0.3 = 尝试性的，0.9 = 近乎确定
+- **领域标签 (Domain-tagged)** — code-style、testing、git、debugging、workflow 等
+- **证据支持 (Evidence-backed)** — 追踪是哪些观察结果创建了它

-## How It Works
+## 工作原理

 ```
-Session Activity
+会话活动 (Session Activity)
      │
-      │ Hooks capture prompts + tool use (100% reliable)
+      │ 钩子捕获提示词 + 工具使用 (100% 可靠)
      ▼
 ┌─────────────────────────────────────────┐
 │         observations.jsonl              │
-│   (prompts, tool calls, outcomes)       │
+│   (提示词, 工具调用, 结果)                 │
 └─────────────────────────────────────────┘
      │
-      │ Observer agent reads (background, Haiku)
+      │ 观察者智能体读取 (后台运行, Haiku)
      ▼
 ┌─────────────────────────────────────────┐
-│          PATTERN DETECTION              │
-│   • User corrections → instinct         │
-│   • Error resolutions → instinct        │
-│   • Repeated workflows → instinct       │
+│          模式检测 (PATTERN DETECTION)    │
+│   • 用户修正 → 直觉                       │
+│   • 错误修复 → 直觉                       │
+│   • 重复工作流 → 直觉                     │
 └─────────────────────────────────────────┘
      │
-      │ Creates/updates
+      │ 创建/更新
      ▼
 ┌─────────────────────────────────────────┐
 │         instincts/personal/             │
@@ -78,7 +78,7 @@ Session Activity
 │   • use-zod-validation.md (0.6)         │
 └─────────────────────────────────────────┘
      │
-      │ /evolve clusters
+      │ /evolve 聚类
      ▼
 ┌─────────────────────────────────────────┐
 │              evolved/                   │
@@ -88,11 +88,11 @@ Session Activity
 └─────────────────────────────────────────┘
 ```

-## Quick Start
+## 快速开始

-### 1. Enable Observation Hooks
+### 1. 启用观察钩子 (Observation Hooks)

-Add to your `~/.claude/settings.json`:
+添加到你的 `~/.claude/settings.json`：

 ```json
 {
@@ -115,34 +115,34 @@ Add to your `~/.claude/settings.json`:
 }
 ```

-### 2. Initialize Directory Structure
+### 2. 初始化目录结构

 ```bash
 mkdir -p ~/.claude/homunculus/{instincts/{personal,inherited},evolved/{agents,skills,commands}}
 touch ~/.claude/homunculus/observations.jsonl
 ```

-### 3. Run the Observer Agent (Optional)
+### 3. 运行观察者智能体 (可选)

-The observer can run in the background analyzing observations:
+观察者可以在后台运行，分析观察结果：

 ```bash
-# Start background observer
+# 启动后台观察者
 ~/.claude/skills/continuous-learning-v2/agents/start-observer.sh
 ```

-## Commands
+## 命令

-| Command | Description |
+| 命令 | 描述 |
 |---------|-------------|
-| `/instinct-status` | Show all learned instincts with confidence |
-| `/evolve` | Cluster related instincts into skills/commands |
-| `/instinct-export` | Export instincts for sharing |
-| `/instinct-import <file>` | Import instincts from others |
+| `/instinct-status` | 显示所有习得的直觉及其置信度 |
+| `/evolve` | 将相关的直觉聚类为技能/命令 |
+| `/instinct-export` | 导出直觉以便共享 |
+| `/instinct-import <file>` | 从他人处导入直觉 |

-## Configuration
+## 配置

-Edit `config.json`:
+编辑 `config.json`：

 ```json
 {
@@ -178,80 +178,80 @@ Edit `config.json`:
 }
 ```

-## File Structure
+## 文件结构

 ```
 ~/.claude/homunculus/
-├── identity.json           # Your profile, technical level
-├── observations.jsonl      # Current session observations
-├── observations.archive/   # Processed observations
+├── identity.json           # 你的个人资料、技术水平
+├── observations.jsonl      # 当前会话观察结果
+├── observations.archive/   # 已处理的观察结果
 ├── instincts/
-│   ├── personal/           # Auto-learned instincts
-│   └── inherited/          # Imported from others
+│   ├── personal/           # 自动习得的直觉
+│   └── inherited/          # 从他人处导入的直觉
 └── evolved/
-    ├── agents/             # Generated specialist agents
-    ├── skills/             # Generated skills
-    └── commands/           # Generated commands
+    ├── agents/             # 生成的专家智能体
+    ├── skills/             # 生成的技能
+    └── commands/           # 生成的命令
 ```

-## Integration with Skill Creator
+## 与技能创建器 (Skill Creator) 集成

-When you use the [Skill Creator GitHub App](https://skill-creator.app), it now generates **both**:
- Traditional SKILL.md files (for backward compatibility)
- Instinct collections (for v2 learning system)
+当你使用 [Skill Creator GitHub App](https://skill-creator.app) 时，它现在会**同时**生成：
+- 传统的 SKILL.md 文件（为了向后兼容）
+- 直觉集合 (Instinct collections)（为了 v2 学习系统）

-Instincts from repo analysis have `source: "repo-analysis"` and include the source repository URL.
+来自仓库分析的直觉带有 `source: "repo-analysis"` 标签，并包含源仓库的 URL。

-## Confidence Scoring
+## 置信度评分 (Confidence Scoring)

-Confidence evolves over time:
+置信度会随时间演化：

-| Score | Meaning | Behavior |
+| 分数 | 含义 | 行为 |
 |-------|---------|----------|
-| 0.3 | Tentative | Suggested but not enforced |
-| 0.5 | Moderate | Applied when relevant |
-| 0.7 | Strong | Auto-approved for application |
-| 0.9 | Near-certain | Core behavior |
+| 0.3 | 尝试性 (Tentative) | 建议但不强制执行 |
+| 0.5 | 中等 (Moderate) | 在相关时应用 |
+| 0.7 | 强 (Strong) | 自动批准应用 |
+| 0.9 | 近乎确定 (Near-certain) | 核心行为 |

-**Confidence increases** when:
- Pattern is repeatedly observed
- User doesn't correct the suggested behavior
- Similar instincts from other sources agree
+**置信度增加** 的情况：
+- 模式被重复观察到
+- 用户没有修正建议的行为
+- 来自其他来源的类似直觉达成一致

-**Confidence decreases** when:
- User explicitly corrects the behavior
- Pattern isn't observed for extended periods
- Contradicting evidence appears
+**置信度降低** 的情况：
+- 用户明确修正了该行为
+- 模式在很长一段时间内没有被观察到
+- 出现矛盾的证据

-## Why Hooks vs Skills for Observation?
+## 为什么使用钩子 (Hooks) 而非技能 (Skills) 进行观察？

-> "v1 relied on skills to observe. Skills are probabilistic—they fire ~50-80% of the time based on Claude's judgment."
+> “v1 依赖技能进行观察。技能是概率性的——根据 Claude 的判断，它们的触发率约为 50-80%。”

-Hooks fire **100% of the time**, deterministically. This means:
- Every tool call is observed
- No patterns are missed
- Learning is comprehensive
+钩子（Hooks）的触发是 **100% 确定性**的。这意味着：
+- 每一个工具调用都会被观察到
+- 不会遗漏任何模式
+- 学习是全面的

-## Backward Compatibility
+## 向后兼容性

-v2 is fully compatible with v1:
- Existing `~/.claude/skills/learned/` skills still work
- Stop hook still runs (but now also feeds into v2)
- Gradual migration path: run both in parallel
+v2 完全兼容 v1：
+- 现有的 `~/.claude/skills/learned/` 技能仍然有效
+- Stop 钩子仍然运行（但现在也会输入到 v2 系统中）
+- 渐进式迁移路径：并行运行两者

-## Privacy
+## 隐私

- Observations stay **local** on your machine
- Only **instincts** (patterns) can be exported
- No actual code or conversation content is shared
- You control what gets exported
+- 观察结果保存在你的本地机器上
+- 只有 **直觉（instincts）**（即模式）可以被导出
+- 不会共享实际的代码或对话内容
+- 你可以控制哪些内容被导出

-## Related
+## 相关链接

- [Skill Creator](https://skill-creator.app) - Generate instincts from repo history
- [Homunculus](https://github.com/humanplane/homunculus) - Inspiration for v2 architecture
- [The Longform Guide](https://x.com/affaanmustafa/status/2014040193557471352) - Continuous learning section
+- [Skill Creator](https://skill-creator.app) - 从仓库历史生成直觉
+- [Homunculus](https://github.com/humanplane/homunculus) - v2 架构的灵感来源
+- [The Longform Guide](https://x.com/affaanmustafa/status/2014040193557471352) - 持续学习章节

 ---

-*Instinct-based learning: teaching Claude your patterns, one observation at a time.*
+*基于直觉的学习：通过一次又一次的观察，教会 Claude 你的模式。*