Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add commands/xiaobei930/cc-best/evolvegit clone --depth 1 https://github.com/xiaobei930/cc-bestWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00014 | $0.02331 |
| Opus 5 | $0.00007 | $0.01166 |
| Sonnet 5 | $0.00003 | $0.00466 |
| Haiku 4.5 | $0.00001 | $0.00233 |
Grade A, and why
evolve scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 332 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/evolve - 知识演化
将积累的学习内容聚类分析，演化为更高级的结构：Skills、Agents 或 Commands。核心原则：量变引发质变，经验升华为能力。
用法
/cc-best:evolve # 分析所有学习内容，建议演化
/cc-best:evolve --execute # 执行演化，生成文件
/cc-best:evolve --domain testing # 仅分析测试领域
/cc-best:evolve --threshold 5 # 要求 5+ 条相关知识才聚类
/cc-best:evolve --dry-run # 预览不执行
角色定位
- 身份: 知识架构师
- 目标: 将碎片化知识升华为系统化能力
- 原则: 模式重复 3 次，就值得抽象
核心理念
当同类知识积累到一定数量，就应该被抽象为更高级的结构
演化规则
演化为 Command（用户主动调用）
触发条件: 多条知识描述的是"用户请求时的操作序列"
示例：
- 知识1: "创建数据库表时，先写 migration"
- 知识2: "创建数据库表后，更新 schema"
- 知识3: "创建数据库表后，重新生成类型"
→ 演化为: /new-table 命令
特征:
- 触发词包含"当用户请求"、"当创建"、"当添加"
- 描述的是可重复的操作序列
- 有明确的起点和终点
演化为 Skill（自动触发的能力）
触发条件: 多条知识描述的是"特定场景下的自动行为"
示例：
- 知识1: "写函数时，优先使用函数式风格"
- 知识2: "修改状态时，使用不可变模式"
- 知识3: "设计模块时，避免类继承"
→ 演化为: functional-patterns Skill
特征:
- 触发词包含"当写代码时"、"当遇到"、"总是"
- 描述的是编码风格或模式偏好
- 应该在背景中持续生效
演化为 Agent（需要深度隔离的复杂任务）
触发条件: 多条知识描述的是"需要多步骤、多工具的复杂流程"
示例：
- 知识1: "调试时，先检查日志"
- 知识2: "调试时，隔离失败组件"
- 知识3: "调试时，创建最小复现"
- 知识4: "调试时，用测试验证修复"
→ 演化为: debugger Agent
特征:
- 描述的是多步骤流程
- 需要多种工具配合
- 适合在独立上下文中执行
演化判定矩阵
| 知识特征 | 演化类型 | 理由 |
|---|---|---|
| 用户明确请求触发 | Command | 需要用户意图驱动 |
| 代码编写时自动应用 | Skill | 持续生效的背景知识 |
| 复杂多步骤流程 | Agent | 需要隔离和深度处理 |
| 简单的单条规则 | Hook | 用于自动检查 |
| 架构级别的约束 | CLAUDE.md | 宪法级别规则 |
工作流程
1. 知识收集
├─ 读取 memory-bank/*.md
├─ 读取 rules/*.md 中的规则
├─ 读取 /cc-best:learn --status 的内容
└─ 统计每条知识的置信度和使用频率
2. 聚类分析
├─ 按领域分组（testing, git, api, style...）
├─ 按触发模式分组
├─ 按行为类型分组
└─ 识别 3+ 条相关知识的聚类
3. 演化判定
├─ 分析聚类的特征
├─ 确定演化类型（Command/Skill/Agent）
├─ 计算演化置信度
└─ 生成演化建议
4. 文件生成
├─ 生成 commands/*.md（Command）
├─ 生成 skills/*/SKILL.md（Skill）
├─ 生成 agents/*.md（Agent）
└─ 标记源知识为"已演化"
5. 关联更新
├─ 更新 COMMANDS.md
├─ 更新 skills/README
├─ 更新 agents/README
└─ 更新 CHANGELOG.md
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 332 lines · 14 tokens per session scan A f60fd2b12ab2
evolve is a command published in the GitHub repository xiaobei930/cc-best (50 stars, last pushed 2mo ago), licensed MIT. It adds 14 tokens to every session and 2,331 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
guide
Interactive guide to fellowship. Walks you through a real task using the structured research-plan-implement flow, then shows you what's next.
revise-claude-md
이번 세션에서 학습한 사실을 Hames 라우팅 매트릭스에 따라 적절한 모듈/워크스페이스 CLAUDE.md 에 반영.
change-quiz
Generate a change report and a must-pass comprehension quiz before merging.
auto-optimize
Self-improving agent optimization — generates challenger variants of any agent/command, benchmarks against baseline, promotes winners, logs learnings to instincts. Inspired by Karpathy's autoresearch pattern.
learn-mode
Interactive code tutor — breaks down codebase logic, explains complexities, translates technical concepts for the user. Ideal after /btw commands. Teaches the WHY behind the code, not just the WHAT.
session-validate
End-of-session self-training — captures session metrics, extracts patterns via metaclaw-learner, updates instincts, and generates optimization hypotheses for the next run.