No red flags found in any of the 11 categories — no credential harvesting, no data exfiltration, no curl-pipe-shell installer. Scanned against the SlowMist agent-security taxonomy, refreshed every 8 hours. Full audit →
by xiaonancs · Agent Tool · ★ 67
Last updated: · Indexed by AgentSkillsHub · Auto-synced every 8h
🔒 Is claude-code-source-analysis safe to install? View the security audit →
Claude Code 源码解析 基于逆向还原的 源码快照，系统级工程视角重建 Claude Code 的完整 Agent Harness 实现。 30 篇文章 / 23 个子系统 / 179 张架构图 / 26,000+ 行研究文本。 系列研究 本研究是 AI Agent / Coding Harness 源码深度研究系列的一部分，姊妹篇： OpenAI Codex 源码深度研究 — OpenAI Codex CLI + Rust workspace + TS/Python SDK 全栈（27 篇 / 51 万字 / 149 张 Mermaid 图） Hermes Agent 深度研究 — Hermes Agent + 4 个同源/对比项目（OpenClaw / EvoMap / OpenHarness / JiuwenClaw），含抄袭分析（38 篇 / 32 章 / 97 张架构图） OpenClaw 源码深度研究 — 358k Stars OpenClaw v2026.4.15 源码 + 23k commits + 生态调研 引言 如果把 Claude Code 只看成"会写代码的命令行工具"，许多设计都会显得过重。如果把它放回到 Agent 运行时系统的语境中，启动链路、上下文管理、权限体系、任务调度、终端渲染与扩展协议之间的关系就会清晰得多。本文采用"结构先于结论"的组织方式：先交代对象、边界与路径，再讨论判断与启示。 整体结构 整体采用"两层前言 + 三部正文"的展开方式。 交代社区讨论的背景、问题意识和阅读本研究的理由。 交代 Claude Code 最重要的技术主线。 处理入口、启动、状态、提示词、主循环、上下文、记忆与缓存。 处理工具、安全、权限、配置、命令、Agent、任务、MCP 与 Hooks。 处理特性开关、API 适配、渲染引擎与设计系统。 收纳扩展点、架构模式、消息附件、LSP 与源码归档。 负责提出问题， 负责给出读法，正文三部则分别展开系统的基础层、执行层与工程层。 参考来源 Claude-Code-Source-Study — 25 篇同主题深度分析，包含详细源码引用和设计模式分析 Harness 技术深度分析 — 基于源码的 Harness 核心技术完整剖析 全网 120+ 篇技术文章调研 — 社区认知的系统性梳理与误解纠正 Anthropic 官方工程博客 — 第一手的设计意图和架构决策 目录 总纲 全网技术文章调研与核心认知 [Harness 技术深度分析](总纲B-Claude-Code-Har
| Stars | 67 |
| Forks | 11 |
| Language | JavaScript |
| Category | Agent Tool |
| Quality Score | 49.9043761226032/100 |
| Last Updated | 2026-05-27 |
| Created | 2026-04-13 |
| Platforms | claude-code, node |
| Est. Tokens | ~1091k |
These tools work well together with claude-code-source-analysis for enhanced workflows:
Looking for a claude-code-source-analysis alternative? If you're comparing claude-code-source-analysis with other agent tool tools, these 6 projects are the closest alternatives on Agent Skills Hub — ranked by topic overlap, star count, and community traction.
Source code teardowns of 14 AI coding agents. What is actually inside Claude Code, Dify, OpenHands, Cline, Mem
Just another desktop client for OpenCode 2. Manage projects, sessions, parallel agents, requests, and changes
Deconstructing Claude Code — interactive source code analysis, guided wc/helpe walkthroughs, and deep architec
Any Agent × Any Provider × Anywhere. Connect Cursor, Claude Code, OpenCode, or Codex to Slack, Telegram, Disco
Portable CC-inspired skills for memory, verification, multi-agent coordination, context compression, and proac
Local-first coordination for human and agent work: durable work, decisions, dispatches, evidence, and prompt-f
Explore other popular agent tool tools:
claude-code-source-analysis is Claude Code 源码深度研究，包括 Foundations/Execution/Infrastructure 三大章节和 23 个子系统的架构分析拆解。. It is categorized as a Agent Tool with 67 GitHub stars.
claude-code-source-analysis is primarily written in JavaScript. It covers topics such as agent-harness, ai-agent, anthropic.
You can find installation instructions and usage details in the claude-code-source-analysis GitHub repository at github.com/xiaonancs/claude-code-source-analysis. The project has 67 stars and 11 forks, indicating an active community.
The top alternatives to claude-code-source-analysis on Agent Skills Hub include awesome-ai-anatomy, palot, claude-harness. Each offers a different approach to the same problem space — compare them side-by-side by stars, quality score, and community activity.
Grades come from a rule-based scan built on the SlowMist agent-security taxonomy, covering 11 red-flag categories including credential harvesting, data exfiltration, and curl | sh installers. It is a first-layer scan, not a manual audit — we say so rather than overstate it.
The scale of the problem is documented independently: Liu et al. (2026), in a study of 31,132 agent skills, report that 26.1% contain security vulnerabilities. Our own full-catalog census is published as a citable open dataset.
Sources & who's responsible: