DeusData/codebase-memory-mcp
High-performance code intelligence MCP server. Indexes codebases into a persistent knowledge graph — average repo in milliseconds. 158 languages, sub-ms queries, 99% fewer tokens. Single static binary, zero dependencies.
C · ★ 44,763 · Fork 3,648 · 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
DeusData/codebase-memory-mcp 已在 GitHub Trending 每日榜单上榜 12 天,最高排名第 1,首次上榜 2026-06-17,最近一次上榜 2026-09-24。
| 日期 | 排名 | Star | 当日增长 |
|---|---|---|---|
| 2026-09-24 | #14 | 44,763 | +190 |
| 2026-09-23 | #14 | 44,516 | +266 |
| 2026-06-29 | #6 | 20,632 | +2,190 |
| 2026-06-28 | #6 | 19,463 | +2,162 |
| 2026-06-24 | #14 | 13,652 | +1,300 |
| 2026-06-23 | #11 | 12,782 | +1,299 |
| 2026-06-22 | #8 | 11,421 | +1,186 |
| 2026-06-21 | #5 | 10,179 | +1,271 |
| 2026-06-20 | #1 | 9,248 | +1,267 |
| 2026-06-19 | #1 | 8,119 | +2,322 |
| 2026-06-18 | #1 | 6,925 | +2,308 |
| 2026-06-17 | #1 | 5,091 | +718 |
同期上榜的项目
- rohitg00/ai-engineering-from-scratch – Learn AI Engineering! Learn it. Build it. Ship it for others.
- vectorize-io/hindsight – Hindsight: Agent Memory That Learns
- dream-num/univer – Build your own AI office with Univer Office SDK. Bring editing, calculation, collaboration, and file processing into your product for people and AI agents. Built on an extensible open-source core.
- google/ax – Agent Executor (AX) is Google's open agentic orchestration runtime
- NVIDIA/Model-Optimizer – A unified library of SOTA model optimization techniques like quantization, distillation, pruning, neural architecture search, speculative decoding, etc. It compresses deep learning models for downstream deployment frameworks like TensorRT-LLM, TensorRT, vLLM, etc. to optimize inference speed.
- FxEmbed/FxEmbed – Fix X/Twitter and Bluesky embeds! Use multiple images, videos, polls, translations and more on Discord, Telegram and others
codebase-memory-mcp 项目深度剖析与商业评估报告
codebase-memory-mcp 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
根本痛点的精准打击 该项目直击AI编码辅助领域最核心的效率瓶颈:代码理解成本。传统AI Agent通过逐文件读取(grep/read)理解代码库,导致大量token浪费(作者实测5个查询节省99.2% token)和上下文窗口污染。codebase-memory-mcp通过构建持久化知识图谱,将代码结构查询从O(n)文件检索降为O(1)图遍历。
企业级商业落地场景 1. 大型代码库的CI/CD管道集成:Linux内核(28M LOC)3分钟索引完成,可嵌入CI流程作为代码审查的前置分析步骤,自动检测变更影响范围 2. 微服务架构治理:跨服务HTTP/gRPC调用链路自动发现,支持多仓库联合索引(CROSS edges),解决企业级服务依赖关系可视化难题 3. 技术债务管理:Louvain社区检测+死代码检测+ADR管理三合一,提供代码库健康度仪表盘 4. 安全审计辅助:本地化运行(100% offline)、零网络依赖,满足金融/医疗等合规要求
竞争差异化与护城河 | 维度 | codebase-memory-mcp | 竞品(Sourcegraph/Cody/CodeQL) | |------|---------------------|-------------------------------| | 部署复杂度 | 单静态二进制,零依赖 | 通常需要Docker/Java运行时 | | 索引速度 | 毫秒级(小库)/分钟级(大库) | 分钟级(小库)/小时级(大库) | | 语言覆盖 | 158语言,含Hybrid LSP类型解析 | 通常<50语言 | | Token效率 | 99%+减少 | 通常仅减少50-70% | | 企业合规 | SLSA Level 3 + VirusTotal + 本地化 | 多数依赖云服务 |
护城河在于:纯C实现+零依赖+树形语法静态编译的极致性能,以及Hybrid LSP层提供的语义级精确度,这是Python/Node.js实现难以复现的技术壁垒。
2. 核心架构设计与技术深度
模块拆解思想 ┌──────────────────────────────────────────────┐ │ main.c │ │ (MCP stdio server + CLI + install/update) │ ├────────┬────────┬────────┬────────┬──────────┤ │ mcp/ │ cli/ │ store/ │ cypher/│ pipeline/│ │ (14工具)│ (11代理)│ (图存储)│ (查询引擎)│ (多通索引)│ ├────────┴────────┴────────┴────────┴──────────┤ │ foundation/ (平台抽象层) │ │ (线程/文件系统/日志/内存管理) │ ├──────────────────────────────────────────────┤ │ internal/cbm/ (158个tree-sitter语法) │ │ + Hybrid LSP (9语言类型解析器) │ └──────────────────────────────────────────────┘
在 README 中展示上榜徽章
[](https://github.fq6825.top/repo/DeusData/codebase-memory-mcp)