
Creator of OpenClaw & Agent Engineer |OpenAI
Creator of OpenClaw (356K+ stars), the open-source AI personal assistant that became one of the fastest-growing projects in GitHub history. Previously founded PSPDFKit (1B+ devices, exited 2021). Joined OpenAI in Feb 2026 to build personal agents for everyone. Self-described 'Clawdfather' who advocates for agentic engineering and owns-your-data AI.
GitHub
34 repositories · 381.9k total stars
Your own personal AI assistant. Any OS. Any Platform. The lobster way.
Skill Directory for OpenClaw
All versions of all skills on clawhub.com archived
Headless CLI client for stateful Agent Client Protocol (ACP) sessions
OpenClaw-native workflow shell: composable pipelines and safe automations
Packages OpenClaw for nix.
Automated, hardened OpenClaw installation with Tailscale VPN, UFW firewall, and Docker isolation
Windows companion suite for OpenClaw - System Tray app, Shared library, Node, and PowerToys extension
Languages
Blog
Announces joining OpenAI to build accessible AI agents while transitioning OpenClaw to an independent foundation.
Details how modern AI models have dramatically accelerated his development workflow, allowing shipping at unprecedented speeds.
Technical deep-dive into a subtle macOS signing bug and the debugging process to resolve it.
Advocates for direct, intuitive interaction with AI models over complex frameworks and elaborate harnesses.
Co-created with Orta Therox, a meetup format for AI-assisted developers: lightning talks, no recordings, no selling.
Live demonstration of building a complete application using AI coding tools.
Definitive guide to his AI development setup: Ghostty + Claude Code + minimal tooling. Context is precious, less is more.
Curated list of must-read resources for developers working with AI coding agents.
Reflection on the addictive nature of AI-assisted coding and the 'just one more prompt' compulsion.
Introduces Poltergeist, a tool for keeping builds fresh and automating build pipeline maintenance.
Critique of AI-generated startup content and generic advice flooding the tech ecosystem.
Monthly curated reading list for the agentic engineering community.
Guide to running AI models locally, covering infrastructure choices and practical tradeoffs.
Investigation into privacy issues in logging frameworks and system-level data collection.
Retrospective on one year of VibeTunnel — turning browsers into Mac terminals for remote AI development.
Complete guide to integrating AppleScript with command-line tools on macOS.
Major update to Peekaboo — lightning-fast macOS screenshots for AI agents, now available as standalone CLI.
Updated guide to running multiple Claude Code instances in parallel for maximum development throughput.
Foundation reading list for developers entering the agentic engineering space.
Raw account of 16-hour AI-assisted coding days and the dopamine-driven loop of rapid iteration.
Workflow for using AI (Gemini) to rapidly understand unfamiliar codebases.
Introduces stats.store, a privacy-first analytics solution for macOS apps using Sparkle.
Technical guide to displaying settings windows from macOS menu bar applications.
Launches VibeTunnel, enabling browser-based terminal access to Mac for remote AI development.
Updated Vibe Meter for tracking Claude Code costs and usage patterns.
Introduces llm.codes for making developer documentation AI-friendly and searchable.
Deep technical exploration of automatic observation tracking across Apple’s UI frameworks.
Introduces Peekaboo, an MCP server enabling AI agents to take and analyze macOS screenshots instantly.
Practical guide to migrating a large test suite from XCTest to the new Swift Testing framework.
Guide to running multiple concurrent Claude Code instances for parallel development — the multi-agent workflow.
War story of macOS code signing and notarization challenges with the Sparkle update framework.
Introduces Vibe Meter, a macOS menu bar app for monitoring AI coding tool spending.
Runs Claude Code with full system permissions, using AI as a universal computer interface for coding, automation, and system management.
Practical advice on choosing AI tool subscriptions without overthinking the decision.
Open-source Swift library for converting HTML to Markdown, useful for AI agent pipelines.
Live workshop building two functional apps in 3 hours, demonstrating AI-assisted development where code becomes cheaper and iteration becomes faster.
Guide to best practices for building and using Model Context Protocol servers with AI coding tools.
After selling PSPDFKit and struggling with post-exit emptiness, rediscovered passion through building with AI.
Technical solution for controlling menu visibility in SwiftUI macOS applications.
Workaround for SwiftUI keyboard shortcut bugs in macOS apps.
Technique for handling simultaneous tap and long-press gestures on SwiftUI buttons.
Early experience report using M1 Mac Minis for CI/CD pipelines.
Comprehensive developer review of Apple’s M1 chip covering performance, compatibility, and development workflow impact.
Strategies for curating a useful Twitter feed by muting noise and following high-signal accounts.
Practical guide to growing a developer-focused Twitter following through authentic engagement.
Critical assessment of SwiftUI’s maturity and readiness for production use.
Investigation into UIKit controls that are silently blocked in Mac Catalyst Mac Idiom mode.
Workaround for SwiftUI’s automatic keyboard avoidance behavior in UIHostingController.
Guide to implementing effective logging in Swift applications using os_log and custom solutions.
How to build and test projects with unreleased Swift compiler snapshots from trunk.
Advanced technique for calling super implementations at runtime in Swift using the Objective-C runtime.
Review of zld, a drop-in faster replacement for Apple’s ld linker for faster build times.
Introduces InterposeKit for elegant method swizzling in Swift, enabling runtime behavior modification.
Detailed investigation into a recurring text input crash in Mac Catalyst applications.
How iOS jailbreaking enables better debugging and understanding of UIKit internals.
Blog relaunch post, committing to regular writing about iOS/macOS development.
Inside look at PSPDFKit’s distributed team culture, processes, and engineering practices.
Advanced techniques for runtime method swizzling in Swift applications.
Practical guide for first-time WWDC attendees covering logistics, networking, and sessions.
Technical challenges encountered implementing drag-and-drop in iOS apps, especially for PDF documents.
Early exploration of Apple’s Marzipan (later Catalyst) for bringing iOS apps to macOS.
Strategies for managing Slack notifications and maintaining focus in distributed teams.
Advanced debugging techniques for iOS/macOS development including LLDB, dtrace, and runtime inspection.
Challenges and solutions for distributing pre-compiled Swift frameworks.
Making Objective-C APIs more Swift-friendly through naming conventions and nullability annotations.
Argument for replacing UITableView with UICollectionView as the default list component in iOS.
Techniques for dramatically speeding up iOS UI test execution.
Lessons from building PSPDFKit’s fully distributed engineering team across multiple countries.
Guide to writing effective Apple Radar bug reports that actually get fixed.
Technical exploration of real-time collaboration in document editing and Apple’s approach.
Best practices for checking iOS versions efficiently in code.
Deep investigation into whether UIImage is thread-safe in practice, with surprising findings.
Techniques for making Objective-C code more idiomatic when used from Swift.
PSPDFKit’s practical approach to cross-platform development across iOS, Android, and web.
Unexpected behaviors and edge cases when using Swift extensions.
How to use ccache to dramatically speed up Objective-C/C++ compilation times.
Analysis of Apple’s ResearchKit framework for medical research applications.
Hidden UIKit debugging features and how to activate them for development.
Introduction to Aspects, a popular iOS library for clean aspect-oriented programming via method swizzling.
Deep dive into safe method swizzling patterns and touch event forwarding in UIKit.
Philosophy of finding and fixing Apple framework bugs through runtime patches — a hallmark of PSPDFKit’s approach.
How to implement UIAppearance protocol support in custom iOS views.
Elegant pattern for intercepting delegate methods using NSProxy in Objective-C.
Announcement of leaving a previous role to focus full-time on PSPDFKit.
Blog reboot — fresh start for writing about iOS development.
Biography
Peter Steinberger is an Austrian software engineer, entrepreneur, and the creator of OpenClaw, one of the fastest-growing open-source projects in GitHub history. Born in rural Upper Austria, he became obsessed with computers at age 14 and went on to study medical computer science at TU Wien, where he launched the university's first Mac and iOS developer course. In 2011, while waiting for a U.S. work visa, he co-founded PSPDFKit with Martin Schurrer, bootstrapping a PDF rendering SDK that grew to power document workflows at Dropbox, Lufthansa, IBM, and over one billion Apple devices. The company raised over EUR 100M from Insight Partners in October 2021 and later rebranded as Nutrient. After exiting PSPDFKit, Steinberger spent three years away from coding, recovering from severe burnout he described as 'complete identity loss, not just fatigue.' In late 2024 he discovered AI-assisted development and called it his 'holy f--- mind-blowing moment.' In November 2025, he built a one-hour prototype connecting WhatsApp to Claude's API that became Clawdbot. The bot autonomously added voice message support by inspecting file headers and routing audio through ffmpeg and Whisper. After a polite trademark concern from Anthropic, the project was renamed MoltBot and then OpenClaw. Within three months it reached over 200,000 GitHub stars and two million visitors in a single week, making it one of the fastest growth curves in GitHub history. On February 14, 2026, Steinberger announced he was joining OpenAI to 'build an agent that even my mum can use,' rejecting offers from Meta where Zuckerberg had reached out personally. OpenClaw transitioned to an independent open-source foundation sponsored by OpenAI. Today the project has over 356,000 GitHub stars and a thriving ecosystem including ClawhHub (a skill directory), Lobster (a workflow shell), and ACPX (an Agent Client Protocol CLI). Steinberger describes himself as the 'Clawdfather' and advocates for 'agentic engineering' over 'vibe coding,' emphasizing that AI replaces repetitive plumbing while developers become architects, editors, and prompt engineers.
Open-source, MIT-licensed AI personal assistant running locally on any OS. Built in TypeScript and Swift, it operates through WhatsApp, Telegram, Discord, and Signal, executing real tasks: email management, browser control, shell commands, workflow scheduling, and PR triage. 356K+ GitHub stars, 72K+ forks, 19K+ org followers. One of the fastest-growing open-source projects in GitHub history.
Community-driven skill directory for OpenClaw, enabling users to discover, share, and install agent skills. The ecosystem has grown to thousands of community-contributed skills archived in the skills repository. 7.9K stars.
Headless CLI client for stateful Agent Client Protocol (ACP) sessions, enabling programmatic interaction with AI agents from the terminal. 2.1K+ stars.
An OpenClaw-native workflow shell: a typed, local-first macro engine that turns skills and tools into composable pipelines and safe automations, letting OpenClaw call workflows in one step. 1.1K+ stars.
CLI and Chrome Extension that summarizes any URL, YouTube video, podcast, or file. Point at content and get the gist. 5.5K+ GitHub stars, Steinberger's most-starred personal repo.
PDF rendering SDK co-founded in 2011, fully bootstrapped for 10 years. Powered document workflows at Dropbox, Lufthansa, IBM, and ran on 1B+ Apple devices. Raised EUR 100M+ from Insight Partners in 2021.
Automated, hardened OpenClaw installation with Tailscale VPN, UFW firewall, and Docker isolation for self-hosting. 557 stars.
NixOS packages and modules for declarative OpenClaw deployment, reflecting the project's strong Nix community. 644 stars.
My next mission is to build an agent that even my mum can use.
I could totally see how OpenClaw could become a huge company. And no, it's not really exciting for me.
Code is becoming an implementation detail... AI agents don't replace developers. They replace repetitive plumbing.
If you wake up in the morning, and you have nothing to look forward to, you have no real challenge, that gets very boring, very fast.
If you can't understand how to run a command line, this is far too dangerous of a project for you to use safely.
Research generated April 13, 2026