We're excited to announce a groundbreaking advancement in SDK generation that solves one of the most persistent challenges in API tooling: choosing between reliability and adaptability.

The Problem We've All Faced

Traditional codegen has served us well—it's fast, reliable, and predictable. Change your API spec and you get identical output every time. But it's also rigid, unable to adapt to the nuances of modern API development or provide intelligent enhancements that developers actually need.

LLM-powered codegen promised to change this with its adaptive intelligence and context-aware capabilities. However, it introduced a new problem: inconsistency. Ask it to generate the same SDK twice, and you'll get different results each time—hardly ideal for production environments.

Our Solution: Hybrid Codegen

Today, we're launching a revolutionary hybrid approach that delivers the best of both worlds. You get the rock-solid reliability of deterministic generation for your core SDK structure, with LLM intelligence strategically layered on top for adaptive features like:

• Intelligent parameter handling that understands context

• Context-aware documentation that actually helps developers

• Smart error recovery that guides users through common issues

• Workflow automation that chains API calls together intelligently

  
  

How Hybrid Codegen Works

The process is elegantly simple yet powerful:

1. Deterministic Foundation: Our system first runs traditional codegen to establish your SDK's core structure—ensuring consistency and reliability across regenerations.

2. AI Enhancement Layer: LLMs then enhance specific components where adaptability adds real value, such as generating contextual examples or creating composite functions that chain multiple API calls together.

3. Surgical Precision: Using structured pattern matching queries (essentially SQL for source code syntax trees), the system precisely targets only the elements that need updating. Instead of overwriting entire files, it surgically modifies specific patterns like function signatures and import statements while preserving all your custom code.

This approach ensures that LLM changes persist across regenerations while maintaining the stability you depend on.

Built for Real Workflows

The LLM component operates through clear rules files that define what the AI can modify and which coding standards to follow. This means you can open your SDKs in popular AI-powered editors like Cursor, Claude Code, Gemini, or GitHub Copilot, and the LLM will intelligently follow your guidelines to enhance the code.

Getting Started Is Simple

Ready to experience hybrid codegen for yourself? Here's how to get started:

Installation

Choose your preferred method:

# npm
npm install -g @sideko/cli
# pip
pip install sideko-py
# homebrew
brew install sideko-inc/tap/sideko
# curl
curl -fsSL https://raw.githubusercontent.com/Sideko-Inc/sideko/main/install.sh | sh

Quick Setup

# Login with social or email
sideko login
# Create your initial SDK
sideko sdk init
# Initialize git for easy change tracking
git init && git add . && git commit -m 'deterministic commit'

Start Building Intelligent Workflows

Once your foundation is set, you can start prompting the AI to create sophisticated workflows. For example:

"create a flight tracking workflow 1. get all flights 2. select the next flight 3. return "enjoy :)" if the flight has in-flight entertainment and ":(" if not"

The AI will generate code that chains these operations together intelligently while maintaining compatibility with your deterministic SDK structure.

Open Source Foundation

We believe in transparency and community collaboration. The core network request libraries that power every SDK are public and open source:

• JavaScript implementation

• Python implementation

• Additional language support open sourcing soon

What's Coming Next

This is just the beginning. In the coming weeks, we'll be rolling out LLM enhancement support for Rust, Go, Java, and C#. We're also expanding our platform with features like automatic OpenAPI-to-SDK sync and enhanced documentation generation.

Join the Community

We'd love to see what you build with hybrid codegen. Join our Slack community to share your projects, ask questions, and connect with other developers pushing the boundaries of API tooling.