zilliztech/claude-context

🚨 BREAKING: AI coding agents can finally understand your entire codebase without exploding your token count. Zilliz just dropped Claude Context an open-source MCP server that makes large projects instantly searchable and usable by any coding agent. No more pasting random files. No more “I don’t have that file in context” replies. Just pure, precise code intelligence. Here’s why developers are going crazy over it: → Semantic + BM25 hybrid search that actually gets your code → AST-based smart chunking (not dumb text splitting) → Incremental Merkle-tree indexing — only updates what changed → ~40% token savings with better retrieval quality → Works with Claude Code, Cursor, Windsurf, VS Code, Gemini, and more → Flexible embeddings (OpenAI, Ollama, Voyage, Gemini) → One-command install and you’re done Tell the agent “find the auth logic” or “show me how payments connect to the DB” and it just works. This is the devtool that turns AI from “sometimes helpful” into “actually knows my project.” 6.2k stars and climbing. Built in TypeScript. 100% Open Source. MIT License.

Matched through a repo-specific project phrase query.

HE BUILT A BOT THAT PREDICTS NBA GAMES BEFORE TIPOFF USING FOUR LAYERS OF SIGNAL. ML model, sportsbook odds, Polymarket pricing, and Claude context all feed into one machine that fires before the first whistle.

Matched through a repo-specific project phrase query.

今日 GitHub Trending 榜首，Claude Context 这个开源项目专门解决 AI 编程工具的上下文窗口不够用的问题。 通过 MCP 插件，让 AI 编程助手拥有对整个代码库的语义搜索能力，精准找到相关代码。 向 AI 提问时只检索相关代码片段，不用把整个项目塞进上下文。 在检索质量相当的前提下，据测试能减少约 40% 的 token 消耗。 GitHub： github.com/zilliztech/claude… 索引过程用了语法树分析来智能切分代码块，还支持增量索引，只处理改动过的文件。 兼容 Claude Code、Cursor、VS Code、Gemini CLI 等十多种主流编程工具。 如果我们经常用 AI 助手处理大型项目，觉得上下文不够用或者 token 花得太多，这个插件值得试试。

Matched through a repo-specific project phrase query.

感觉大家现在对于代码库检索各执一词，最近看了zilliz新开源了一个claude-context，他们就是用RAG做的，主要设计还是在保持语意完整上，因为grep单纯靠字面意思匹配，对token消耗很大。 github.com/zilliztech/claude…

Returned by a high-confidence repo query and contains a visible project phrase, but the exact URL, slug, or package name was not visible.

10 github repos to cut your claude code tokens by 60-90%. rtk, context mode, code-review-graph, token savior, caveman, token-optimizer-mcp, claude-context, and more. all verified, all free. plus which 2-3 to actually stack. aiia.ro/blog/claude-code-tok…

Matched through a repo-specific project phrase query.

ありがとうございます。ClaudeCodeとCursor両方同時に課金は現状苦しいので来月あたりに切り替えてみます。ClaudeCodeってRAGしてないんですね、何となく内部でRAGやってくれてると思いこんでました。一旦、claude-contextを試して見ようかと思います。 github.com/zilliztech/claude…

Returned by a high-confidence repo query and contains a visible project phrase, but the exact URL, slug, or package name was not visible.

As I understand it the benefit of a harness is being able to own your personal context and jump from model to model. W/o it you have your Claude context, your codex context, your Gemini context etc. Each chat has to learn your workflows from scratch. A harness bridges them all.

Matched through a repo-specific project phrase query.

I just stumbled upon an AI coding tool that transforms how we handle code context. It's called Claude Context. Here’s the scoop: 🧵

Matched through a repo-specific project phrase query.

I've had good results using Claude Context with Codex: github.com/zilliztech/claude… I run the vectordb and embedding model locally. I've barely seen Codex do file searches after implementing it, and it finds important context easily.

Returned by a high-confidence repo query and contains a visible project phrase, but the exact URL, slug, or package name was not visible.