OpenClaw creator working on agents at OpenAI
Peter Steinberger
Peter Steinberger created OpenClaw and works on AI agents at OpenAI. He previously founded PSPDFKit and writes about AI-assisted software development.
About Peter Steinberger
Peter Steinberger is a software developer who created OpenClaw and works on AI agents at OpenAI. He announced joining OpenAI in February 2026, describing his goal as making agents accessible to more people while keeping OpenClaw open and independent.
He previously founded PSPDFKit, a software toolkit for working with PDF documents. His public projects and technical writing cover AI coding workflows, developer tools, and software for macOS and Swift. He publishes under the username steipete.
Mentioned in our coverage
35 articles name Peter Steinberger, excluding author credits.
Page 1 of 3
- OpenClaw Deleted 400K Lines of AI-Written Tests. The test-audit Skill Shows How →
Peter Steinberger says OpenClaw deleted roughly 400,000 lines of AI-generated tests without much change in code coverage, using a test-audit skill that is now public in the repo. The skill is a reusable authoring gate plus an audit workflow, and its rules apply to any codebase where agents write tests.
- What Is Soul.md? Meta Muse's AI Persona File, Explained →
A single viral tweet from OpenClaw creator Peter Steinberger — "I saw their Soul.md file and now i'm curious" — turned a Meta Muse invite code into a case study on the fastest-growing convention in AI tooling: giving an agent its identity in a plain markdown file. Here's what's actually confirmed about Soul.md, and how it fits alongside CLAUDE.md and AGENTS.md.
- Meta Launches Muse: The Personal Agent With a Sentinel Security Architecture →
Meta shipped Muse on September 8-9, 2026 — a 24/7 personal agent for iOS, Android, web, and WhatsApp, built on Muse Spark 1.3. What makes it worth a deep read isn't the assistant pitch, it's the security architecture behind it: a per-user Secure VM, a Sentinel agent that brokers every network request, eBPF-based taint tracking, and a public bug bounty paying up to $130,000 for a working prompt injection.
- Meta's "Hatch" AI Agent Runs on Claude While Meta Spends $10B on Anthropic →
Meta is building a consumer AI agent called "Hatch" that currently runs on Anthropic's Claude models — and, per the New York Times, is separately projecting up to $10 billion a year in spending on Anthropic's AI tools. Here's what's confirmed, what's still reported-not-official, and why even Meta hedges with a competitor's models.