Build
Give the job to an agent
Use LP1 Agent with your own OpenRouter key, or connect Codex, Claude Code, Cursor, or a supported ChatGPT workspace app through LP1 MCP.
LP1 capabilities
LP1 keeps Page Sets, exact first-party measurement, outlier evidence, version-based Experiments, guarded automation, Forms, signed webhooks, Preview, and Publish in one operating surface.
From Page request to next experiment
Build
Use LP1 Agent with your own OpenRouter key, or connect Codex, Claude Code, Cursor, or a supported ChatGPT workspace app through LP1 MCP.
Learn
LP1 keeps views, form impressions, conversions, experiment arms, and outlier evidence together so the next move is obvious.
Operate
Let the agent observe, prepare drafts, start tests, or publish qualified winners inside budgets, cooldowns, stop-loss rules, and explicit owner approval.
Platform support
LP1 exposes the operational pieces agents need, while the app keeps Page Owners focused on Pages, performance evidence, Experiment decisions, and consequential approvals.
Start buildingGroup related Pages and see exact first-party views, form impressions, managed-form conversions, and named custom events.
Use immutable Page Versions as arms, keep visitor assignment stable, and stop shared caching from leaking one arm into another.
LP1 waits for a minimum sample, compares each Page to its set, and explains why a result is—or is not—ready for attention.
Observe, draft, test, and opt-in publish modes keep agent actions inside daily budgets, cooldowns, stop-losses, and scoped token permissions.
Hosting, branded domains, managed forms, GTM, signed webhooks, retries, delivery history, and replay live in the same operating surface.
Read Pages, write Pages, publish, read marketing data, configure operations, run automation, and automate publishing are separate capabilities.