Ship report: April 2026.

This is the first monthly Axint ship report.

The point is simple: show what actually changed, what got cleaner, and what still needs work.

What shipped

• public version and test numbers now stay aligned across the site instead of being hand-copied into multiple places
• PyPI is now on the current version and points Python users to the docs site
• the public install surfaces got tighter across the site, docs, and remote MCP story
• Cloud messaging got sharper around the real workflow: generate, validate, decide, ship
• the benchmark page now reads more like workflow proof and less like a token gimmick

What changed in the public product story

The biggest non-code improvement is that Cloud now reads more like a real product and less like an ambiguous beta.

That matters because the product story needs to be legible fast:

• the compiler is the wedge
• Cloud is the place teams review generated Apple output and shipping risk
• the examples, docs, and integrations need to make that obvious in a few clicks

What got cleaner in distribution

The distribution perimeter was one of the weakest parts of the public surface.

So the work here has been pragmatic:

• fix stale install claims
• tighten integration docs
• make Codex feel first-class instead of hidden
• keep the remote MCP endpoint and public version story aligned

What is still open

Some of the remaining work is still external-surface work:

• directory listings that need to be refreshed or claimed
• more visual proof on integrations and examples
• more third-party signal around usage, stars, and outside conversation

What matters next

The next unlock is outside signal:

• more team conversations
• sharper case studies
• more visible GitHub signal
• real traffic that turns into feedback, stars, and conversations

The product is increasingly coherent.

The job now is to make the outside world feel that coherence too.