Skip to main content
Glama
deenesjakoruzh
Rajahn

RepoWeaver

by Rajahn
OverviewSchemaRelated ServersScoreDiscussions
Python
Local

RepoWeaver

面向 AI 编码代理的代码上下文织物 — 确定性调用图索引、混合检索和单一工具 MCP 接口。

RepoWeaver 构建了一个本地、零成本的代码智能层，让 AI 编码代理（Codex、Claude Code、CodeWiz 等）无需读取整个文件即可回答结构性问题。

"Who calls this method, and what would break if I change its signature?"
→ explore(query, task="impact", repo=".", max_tokens=4000)
→ [verbatim source slices + call paths + blast radius + known blind spots]

为什么

依赖 grep 或全文读取的 AI 代理会浪费 token，并遗漏跨文件调用链。 现有工具要么需要云 API，要么在索引层嵌入 LLM，要么带有限制性许可证。 RepoWeaver 只从该领域中采纳已验证的共识模式，并在本地组装它们。

Related MCP server: codemap

设计

六个原语，均已在多个开源工具中得到独立验证：

#

原语

验证方

1

tree-sitter 确定性解析，本地 $0

CodeGraph · GitNexus · Graft · Aider

2

调用 / 继承 / 导入边 + 爆炸半径

CodeGraph · GitNexus · Graft · CodeQL

3

BM25/FTS + PageRank 图扩散

Graft · CodeGraph · Aider

4

内容哈希增量新鲜度

Graft · Codebase-Memory · CodeGraph

5

MCP 单一强工具 explore()

CodeGraph（经验验证）

6

AGENTS.md 协议注入 + 编辑钩子

GitNexus · CodeGraph · Graft · Potpie

不包括：索引层中的 LLM（所有严肃工具均拒绝）、发送到外部 API 的代码、主观语义摘要。

路线图

里程碑

标签

状态

M1 — Fabric MVP（Java 解析器 + 边 + FTS5 + PageRank + MCP）

v0.1.0

🔜

M2 — 新鲜度与置信度（自动同步 + 消歧 + 置信度边）

v0.2.0

计划中

M3 — 类型精度覆盖层（SCIP/jdtls）

v0.3.0

计划中

M4 — 运行时覆盖层（OTel/Jaeger 追踪 → 边权重）

v0.4.0

计划中

每个里程碑都附带一个在 CI 中运行的 fabric verify --level mN 门禁。

验证方法

基准测试仅使用公共仓库 — 没有专有代码进入此仓库。

基准仓库

目的

spring-projects/spring-petclinic

Spring 依赖注入密集，测试注入感知边

google/gson

与 CodeGraph 公布的 93.3% Java 覆盖率相当

square/okhttp

调用链深度

mybatis/mybatis-3

生成代码噪声过滤

安装（即将在 v0.1.0 中推出）

pip install repoweaver
fabric build          # index current repo, $0, no LLM
fabric init           # inject AGENTS.md protocol + edit hook
fabric verify --level m1

MCP 工具

explore(
  query: str,
  task: "understand" | "impact" | "locate" | "debug",
  repo: str = ".",
  max_tokens: int = 4000
) → {
  slices: [{file, span, source}],
  call_paths: [...],
  blast_radius: [...],
  blind_spots: "DI / reflection / MQ edges not represented"
}

致谢

设计模式源自（并非派生自）： CodeGraph · Graft · GitNexus · Serena · SCIP · Aider

许可证

MIT

This server cannot be installed

RayrunBuild, deploy, and operate MCP servers without becoming an infrastructure team.ray.runAd
A
license - permissive license
Not graded
quality - not tested
A
maintenance

How are these scores calculated?

Maintenance

0dRelease cycle
8Releases (12mo)
Commit activity

Resources

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Connectors

Related MCP Servers

  • codeweave-mcp

    semihkayan
    A
    license
    Not graded
    quality
    D
    maintenance
    An MCP server that gives AI agents structured code understanding and precise code intelligence via local indexing of AST, call graphs, and semantic search.
    76
    4
    Apache 2.0

  • codemap

    abdul-hamid-achik
    A
    license
    Not graded
    quality
    A
    maintenance
    MCP server for local-first code intelligence, providing structural code graph, semantic search, and impact analysis to AI agents.
    2
    MIT

  • uacos

    caotiensinh
    A
    license
    Not graded
    quality
    B
    maintenance
    Local-first code intelligence and safety layer for AI coding agents. MCP server exposes dependency graph, impact analysis, and AST-compressed repo context, backed by typed local memory, patch-scope safety gates, and git-independent transaction rollback.
    1
    MIT

  • LAIN-mcp

    spuentesp
    A
    license
    Not graded
    quality
    A
    maintenance
    A persistent code-intelligence MCP server that builds a queryable knowledge graph of your codebase, enabling AI assistants to perform cross-file structural reasoning, dependency analysis, and blast radius detection.
    6
    MIT

View all related MCP servers

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/Rajahn/code-context-fabric'

If you have feedback or need assistance with the MCP directory API, please join our Discord server