Under Testing · SDK Module

One Map. The Whole Codebase.

LLMs can't fit a complex codebase in context — so they guess file by file, on outdated versions. CDX maps your entire repository and the exact dependency versions you run — eliminating hallucination with end-to-end visibility through call trees across dependencies.

Join the Early Access List Request the Investor Brief

Part of the AIvantGuard SDK — tier-based licensing, usable standalone or as an MCP service.

The problem CDX solves

Context limit

Agents search private repositories file by file, losing the structure that connects files. The big picture is never in view.

Version drift

Models are trained on a snapshot of a library — but your project runs a different version. Interfaces drift; models hallucinate and burn time in trial-and-error.

The CDX answer

A map of the entire codebase with the exact versions of every dependency — AI works with complete structure and the interfaces you actually have. Heterogeneous platforms? CDX fuses their code into the best solution.

Capabilities

Repository-level representation

One map of the whole codebase — not file-by-file search.

Per-crate version experts

456 experts in an MoE architecture, each trained on a specific crate version.

CVE-safe generation

Negative-weight training blocks known vulnerability patterns.

cdx_audit

Scan for vulnerability patterns before you commit.

Behavioral code comparison

Same behavior in different languages, detected automatically.

Version-aware routing

Each dependency matched to the expert trained on your version.

MCP for AI agents

12+ tools, stdio JSON-RPC — works with Claude Code, OpenCode, Codex.

Cross-platform fusion

Composes the best solution from what each platform does well.

Rust native

Zero Python in production. Candle + tree-sitter. Cross-language concept search across 6 languages.

Traction

456

Trained experts

12+

MCP tools

6

Languages

100%

Rust — zero Python in production

VersionMilestoneMeaning
v0.1FoundationPipeline, 8 languages, model scaling
v0.3Agent readyMCP server — agents call it as a tool
v3.0SecurityCVE-aware generation in version experts
v5.0Rust nativeZero Python in production
v6.0Cross-languageTauri IPC trace, concept search, 6 languages

Roadmap

Now

Under testing

CLI and MCP server running on real repositories in daily use. Core engine, experts and security pipeline live.

Next

Product packaging

Wrapping engine, experts and MCP into a distributable product — installers, packaging, release tooling.

Then

Distribution & ecosystem

Integrations with agent platforms, CI pipelines and IDEs — wherever code is written.

Long term

Commercial availability

General availability. Today the engine is under testing — not yet on sale.

Who it is for

AI agent platforms

Repository context through MCP — generation that matches the codebase.

Security teams

Vulnerability-pattern scanning before commit, CVE-aware generation.

Dev tooling

Generation, refactoring candidates, duplication detection across languages.

Migration teams

Behavior-true cross-language migration — confidence, not guesswork.

Give every agent the whole codebase