grft-dev/graftcode

Connect any code. Call as dependency. The official repository for Graftcode: the integration-free development layer that lets you call methods across languages and runtimes on remote services as if they were local. Eliminate integration layer code entirely.

What it solves

Graftcode removes the need to write manual integration layers, SDKs, and client code when connecting different services. It eliminates the overhead of maintaining REST, gRPC, or Thrift APIs, DTOs, and OpenAPI specifications, allowing developers to call remote methods across different languages and runtimes as if they were local dependencies.

How it works

  1. Host: Developers write public methods in their preferred language and run them on the Graftcode Gateway, a multi-runtime host.
  2. Consume: The framework auto-generates strongly-typed clients called "Grafts" which are installed via standard package managers (like npm, pip, or dotnet).
  3. Call: Remote methods are called as native dependencies in the client project, with the framework handling the communication and versioning.
  4. AI Integration: It provides native support for the Model Context Protocol (MCP), turning public static methods into tools for AI agents.

Who it’s for

Software engineers building polyglot systems, microservices, or modular monoliths who want to eliminate integration boilerplate and reduce codebase volume.

Highlights

  • Cross-language support: Works across 20+ languages and 10+ package managers.
  • Zero-code integration: No controllers, IDLs, or client code required.
  • MCP-compatible: Makes backends immediately available as tools for AI agents.
  • Infrastructure agnostic: Supports switching between communication channels (e.g., WebSocket, Kafka, RabbitMQ) without changing code.
  • Efficiency: Claims up to 70% faster performance and lower CPU usage than REST/gRPC.

Related

  • Project
    getsentry/toolkit

    An MCP server that connects Sentry's error and performance data to AI coding agents, enabling natural language querying and debugging within development tools.

  • Project
    yongliang-wu/DFT

    A method for improving LLM generalization during Supervised Fine-Tuning by dynamically rescaling token loss based on predicted probabilities, particularly effective for reasoning tasks.

  • Project
    FlameskyDexive/Legends-Of-Heroes

    A production-ready multiplayer game framework using Unity and .NET 10 that provides server-authoritative state synchronization, an ECS-based combat system, and a hot-update pipeline.

  • Project
    SciML/DiffEqFlux.jl

    A Julia package that integrates differential equation solvers into neural networks to enable the creation of physically informed scientific machine learning models.