Have a Claude agent create and publish a landing page from an AI written brief.
Launch a cross-domain A/B test comparing two landing pages or external URLs.
Track conversions back to the exact page, variant, and channel that produced them.
Pull a CRO performance report through the API for a recurring client check-in.
| alexpilotto/uxon-ai | 0311119/free_registertool | 18597990650-lab/multi-agent-game | |
|---|---|---|---|
| Stars | 24 | 24 | 24 |
| Language | — | Python | Python |
| Setup difficulty | easy | hard | moderate |
| Complexity | 2/5 | 4/5 | 3/5 |
| Audience | pm founder | developer | developer |
Figures from each repo's GitHub metadata at analysis time.
Requires a UXON API key from the hosted platform, the repository itself is documentation, not runnable source code.
UXON AI is a documentation repository for a hosted marketing platform called UXON, which handles landing pages, split testing, and tracking who converts into a customer after clicking an ad. The actual UXON application runs on its own website and is not open source. What lives in this repository instead is a public playbook of commands, examples, and guides meant to let developers and AI agents like Claude control that hosted platform through an API and a connector protocol called MCP. Through these commands, an agent or developer can create blank landing pages, generate one using an AI written brief, or bring in a page built elsewhere with tools like Claude Code or Cursor, then publish it. They can launch A/B experiments, comparing different versions of a page or even entirely different external URLs to see which converts better, including tests that span more than one domain. Conversions themselves get tracked back to the specific page, variant, and marketing channel that produced them, with support for wiring up Google Tag Manager or a direct Google Analytics and Meta setup. The same commands can adjust a client's saved branding, such as colors, logo, and fonts, and pull back reports on how pages and experiments are performing. The repository organizes its guidance by topic, with separate reference pages for landing pages, experiments, conversion tracking, branding, domains, goals, reports, and contacts, plus a set of runbooks that walk through common sequences such as onboarding a new client or running a weekly optimization check. Every command is sent as a JSON object naming the action and its inputs, either directly to an API endpoint using an API key, or through the MCP connector for Claude. The README gives explicit advice for agents using these commands: look up the correct account and resource identifiers before making changes, send only the minimum needed information, keep any high impact action easy to confirm, and read back a resource after changing it to be sure it worked. This is documentation only, so there is no install step or license for a codebase, since the underlying software is a paid, hosted service rather than something you run yourself.
Public documentation and command playbook that lets developers and AI agents control the hosted UXON platform for landing pages, A/B tests, and conversion tracking.
Setup difficulty is rated easy, with roughly 30min to a first successful run.
Mainly pm founder.
This repo across BitVibe Labs
double-check against the repo, no cap.