morluto/rea
Reverse engineer anything with agents, from app behavior down to native binaries.
TypeScript · ★ 460,975 · Fork 112,931 · 在 GitHub 查看
在手机上浏览源码、订阅每日 AI 趋势日报与项目架构深度解读:获取 CopoHub 客户端(iOS & HarmonyOS)→
GitHub Trending 上榜记录
morluto/rea 已在 GitHub Trending 每日榜单上榜 11 天,最高排名第 1,首次上榜 2026-10-06,最近一次上榜 2026-10-16。
| 日期 | 排名 | Star | 当日增长 |
|---|---|---|---|
| 2026-10-16 | #1 | 281,147 | +32,314 |
| 2026-10-15 | #1 | 268,811 | +41,034 |
| 2026-10-14 | #1 | 239,572 | +54,679 |
| 2026-10-13 | #1 | 204,408 | +54,706 |
| 2026-10-12 | #1 | 153,023 | +45,495 |
| 2026-10-11 | #1 | 101,003 | +25,793 |
| 2026-10-10 | #1 | 70,238 | +25,784 |
| 2026-10-09 | #1 | 43,694 | +15,335 |
| 2026-10-08 | #1 | 24,694 | +7,744 |
| 2026-10-07 | #1 | 14,535 | +4,666 |
| 2026-10-06 | #8 | 8,885 | +2,963 |
同期上榜的项目
- anthropics/claude-plugins-official – Official, Anthropic-managed directory of high quality Claude Code Plugins.
- K-Dense-AI/scientific-agent-skills – Turn any AI agent into an AI Scientist. The #1 Agent Skills library for science, used by 250,000+ scientists worldwide. 177 ready-to-use validated skills plus 100+ scientific databases covering biology, chemistry, medicine, and drug discovery. Compatible with Cursor, Claude Code, Codex, Pi, Antigravity, and the open Agent Skills standard.
- affaan-m/ECC – The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
- HKUDS/CLI-Anything – "CLI-Anything: Making ALL Software Agent-Native" -- CLI-Hub: https://clianything.cc/
- mksglu/context-mode – Free, local plugin that keeps tool output out of your coding agent's context window. Sandbox and SQLite FTS5 search, 99% fewer bytes in context. Works with Claude Code, Codex, Pi, OpenCode, Gemini CLI, Copilot CLI and 4 more agents. Zero telemetry.
- ruvnet/ruflo – 🌊 The original agent harness. Deploy intelligent multi-player swarms, coordinate autonomous workflows, and build conversational AI systems. Features adaptive memory, self-learning intelligence, federation, vector RAG integration, and native Claude Code / Codex / Hermes and many more Integrated
REA (morluto/rea) 项目深度剖析与商业评估报告
REA (morluto/rea) 项目深度剖析与商业评估报告
1. 核心定位与商业落地价值
REA 解决的根本痛点并非"反编译"本身,而是将逆向工程从人类专家的手工技艺,转化为可由 AI Agent 编排的、可审计的自动化调查流程。传统工具链(Hopper、IDA、Ghidra)面向人类交互设计,输出的是 GUI 状态与直觉性判断;而 REA 将"打开二进制 → 定位字符串 → 追踪交叉引用 → 反编译 → 重建控制流 → 生成代码"这一链条封装为 MCP 工具目录(Native inspection 41 个、Investigation workflows 14 个、Artifact graph 5 个等),使 Claude Code、Codex、Cursor 等 Agent 能够以结构化方式消费逆向结果。
企业级商业场景: 竞品功能逆向与合规对标:产品团队可要求 Agent 调查某 App 的离线搜索/认证/更新流程,输出证据链后在本项目栈内重建(如 README 中 Notes.app + TypeScript/SQLite 的示例)。 遗留系统与无源码资产接管:企业并购、供应商停止维护时,通过 .NET 程序集元数据、CIL 指令、native 依赖比对快速建立资产认知。 安全审计与供应链验证:Mach-O 代码签名、plist、ASAR 完整性(含 .asar.unpacked SHA-256 校验)、受控进程行为捕获(Process Capture)可用于验证第三方分发物是否被篡改。 CTF 与安全研究:topics 中显式包含 ctf,说明其面向攻防研究场景。
差异化护城河:与 Ghidra/IDA 的 MCP 封装类项目相比,REA 的核心竞争力在于 Evidence(证据)契约——每次结论都携带 artifact 身份、provider、位置、置信度与局限性,且明确区分"观察"与"推断"(如导入的反编译输出被标注为 analyst inference)。这一设计直接回应了 LLM 幻觉问题,是其在 Agent 生态中难以被简单复制的抽象层。
2. 核心架构设计与技术深度
REA 采用清晰的分层 + 注册表驱动架构,Mermaid 流程图揭示了关键链路:
Agent/Terminal → REA(CLI+MCP) → Target-bound Session Router ↓ Deep-provider Registry (确定性选择) ┌───────────┼───────────┬──────────┐ Hopper Ghidra Native macOS Artifact/Browser/Process └───────────┴───────────┴──────────┘ ↓ Owned Provider Runtime (deadline + cleanup)
在 README 中展示上榜徽章
[](https://github.fq6825.top/repo/morluto/rea)