Built by Agents

Meet the OpenClawCrew Team

This entire website, every line of code, every guide, every image, and every marketing asset, was built by a team of AI agents running on OpenClaw. No human wrote a single line of code.

One human (Sameer) gave the direction. The agents did everything else.

The Team

Six specialists, one ship-it mindset.

Anton avatar

Anton

CEO / Project Manager

Model: Claude Opus

What he does: Anton decomposes tasks, assigns work to the right agent, reviews output, enforces quality gates, and makes decisions. He never writes code or docs himself. He delegates, monitors, and ships.

Fun fact: Anton ran this entire build as a 24-hour sprint, coordinating 5 agents across 30+ subagent sessions, with heartbeat checks every 15 minutes.

Marv avatar

Marv

Full-Stack Developer

Model: GPT-5.3 Codex

What he does: Marv builds everything. The Next.js app, Stripe integration, Convex backend, Playwright tests, all of it. He writes code using TDD (test-driven development) and ships with green tests or not at all.

Fun fact: Marv fixed a critical Stripe webhook idempotency bug by refactoring to atomic Convex mutations, all in under 4 minutes.

Kevin avatar

Kevin

Quality Assurance Engineer

Model: GPT-5.2

What he does: Kevin breaks things before users do. He runs Playwright end-to-end tests, audits code for security issues, tests every route, and files bug reports with exact repro steps. Nothing ships without Kevin's sign-off.

Fun fact: Kevin caught a P0 where the Stripe webhook handler used an in-memory idempotency store that would fail on serverless cold starts. Saved us from double-charging customers.

Harry avatar

Harry

Research & Marketing

Model: GPT-5.2

What he does: Harry researches competitors, writes launch copy, creates social media posts, identifies SEO keywords, and drafts content. He also wrote all 10 ultimate guides on this site.

Fun fact: Harry produced a complete launch kit (X thread, Reddit post, LinkedIn post, Indie Hackers post, email announcement, and SEO keyword map) in 55 seconds.

Dexter avatar

Dexter

DevOps & Infrastructure

Model: GPT-5 Mini

What he does: Dexter handles deployment, DNS verification, Vercel configuration, environment variables, and monitoring. He runs hourly deploy-readiness checks and only escalates when something actually needs human action.

Fun fact: Dexter autonomously set up the GitHub repo, linked it to Vercel, and ran deployment verification, all without being asked twice.

Penny avatar

Penny

Docs & Strategy

Model: Claude Sonnet

What he does: Penny writes PRDs, product reviews, gap analyses, and content strategy. She scored every runbook 1-10 and identified exactly what needed improvement to justify the $49 price point.

Fun fact: Penny's product review identified that "production-grade" was an overclaim before any customer could complain, saving us from a trust issue on launch day.

How It Works

This is OpenClaw in action. One orchestrator (Anton) coordinating specialized agents, each with their own model, workspace, and expertise. The same system you can set up with the Starter Kit.

The difference between this team and a human team? They work 24/7, never forget context, and can spin up a new specialist in seconds.

Want to build your own agent crew? The OpenClawCrew Starter Kit gives you the runbooks and templates to get started: openclawcrew.com/checkout