Introduction

Code Storage is a versioned, programmable storage layer that powers the largest agentic workloads in the industry.

Code Storage provides a scalable, distributed storage system built on Git. Its API ergonomics are optimized for agents and machines. No rate limits, complicated auth-flows, or other limitations. And it works whereever your agents work.

What you can do

Work with remote Git repositories through the SDK, the HTTP API, or Git over HTTPS.

• Create repositories, write commits, and merge branches from agent sessions.
• Clone, push, and pull with standard Git.
• Create isolated refs for previews, CI artifacts, and experiments.
• Bulk import existing repositories.
• Sync repositories from GitHub and other hosts into Code Storage.
• Stream files, grep codebases, select globs, and more.

Use cases

Combine our platform’s features to accomplish agent workflows:

• Agent Memory: Save and load agent memory files.
• Capture Agent State: Save intermediate agent work.
• Show Live Diffs: Isolate competing agent attempts.
• Run Parallel Attempts: Isolate competing agent attempts.

How to access

Sign up for an account at code.storage.

View our pricing page for a detailed pricing breakdown.

How it works

Code Storage is a high-scale, distributed storage system built on Git for storing and collaborating on repos, with API ergonomics optimized for machines.

Code Storage leverages Git’s three-phase commit protocol and a “spoke-like” quorum-based architecture to guarantee the durability and consistency of all reads and writes at scale.

Organizations group your resources, registered public keys, and team members. Authentication lets each client make requests with a signed token and the permissions it needs.