Build with agents

Work with agents inside the same folders and app that owns design, content, and assets. The agent reads your real project, proposes edits, and waits for your review.

One place to rule them all

Built in tools and skills

We have built extensive tooling that enables agents to efficiently created, edit and review your site.

Staged, reviewable diffs

Users will visually see whats changed, and decide to keep it or not

Bring your own key

We dont want to be a middleman, bring your own key for Anthropic, Open AI and OpenRouter.

Chat lives with the site

Chat history stored with the site, building on the context of the AI.

AI without a workflow becomes chaos

AI without a workflow becomes chaos

Prompting a coding tool can ship a first page. Ongoing marketing work needs review, constraints, and a place where changes become publishable history.

The agent should see the same files you ship

Guessing from a pasted snippet creates drift. Grounded edits on the live project keep the site coherent.

You should not pay rent on someone else's model meter

Splitleaf does not insert itself as a middleman tax on every token. You bring the key. You pay the provider. The agent stays a feature, not a new subscription layer.

Safe by construction

Project-scoped assistant

The agent works on the currently open site. File access stays sandboxed to that site root.

Safe by construction

Edits are staged with baselines. Writes use precise replacements. Nothing is committed to Git by the agent. You review before publish.

BYOK

Connect your model API key in site settings. Keys are stored for that site's agent connection. Splitleaf is not the LLM bill.

Local-first chat trail

Sessions and staging live under .opt-one/agent/ in the site. Usage can follow the site across devices. The conversation stays near the work.

From install to your first AI-assisted edit

1

Download the app

Install Splitleaf on your machine and sign up free, no credit card, ready in minutes.

2

Design and build for free

Every feature is unlocked while you try it out, the designer, the CMS, the AI agent.

3

Connect to Github and Vercel

Click and connect with a few clicks to store the project in github, and puhblish it on your own domain with vercel.

4

Review the staged diff, then publish

Keep what's right, undo what isn't, and publish through the normal review flow whenever you're ready.

Common questions

Does the agent publish for me?

No. It stages edits. Publish stays a deliberate step with a change review.

Do I need to be technical to use it?

No. The agent is built for non-technical users working on a real site, with reviewable results instead of raw repo commands.

Is agent usage included in the license?

The desktop license covers the app. Model usage is billed by your provider through your key.

Can we keep using external AI tools too?

Yes. Use them for exploration if you want. Use Splitleaf when the change needs to land in the maintained site.

Let AI draft. You decide what ships.

Keep AI speed with a human still on the merge button.

Launching soon