Skip to content

README.md

Latest commit

 

History

History
30 lines (20 loc) · 2.28 KB

README.md

File metadata and controls

30 lines (20 loc) · 2.28 KB

Hey, I'm EJ

I build tools for people running AI agents all day.

Independent freelancer. I work across CTV ads, creative production, and AI-operator tooling. The stuff I open-source comes straight out of my own daily workflow.

What I'm shipping

  • ★ Slash (new): An AI-native, local-first web browser. The assistant can search, read pages, and act on the web for you, running on your own machine with your key or a free local CLI. No account, no telemetry, ads and trackers blocked, passwords and keys sealed with the OS keystore. Open source, so the privacy claims are checkable, not promises. MIT.
  • War Room: Local-first Discord-style cockpit for solo operators running many clients and many AI agents at once. Plug in any OpenAI-Chat-Completions-compatible runtime. AGPL-3.0.
  • OpenWar: Discipline layer for CLI agents. Runs Claude Code, Codex, and Gemini CLI under one phase-gated runtime with deterministic detectors, replayable traces, and persistent memory. MIT.
  • Squire: General-purpose runtime for spawning CLI AI agents as subprocesses with structured event streaming, MCP tool forwarding, and Claude Code permission auto-setup. MIT.
  • api-discover: Point at any website, get a typed API client plus an upfront verdict on whether HTTP-only replay will work (Turnstile, CSRF, WebSocket-delivered results, etc.). MIT.

More tooling across agent ops, video, and motion lives in my repos. When something earns its keep daily, it eventually ships here.

How I work

Small number of clients at a time across creative, video, and AI-tooling work. Operator-first, ship-first. If I'm using it every day, eventually it lands in one of these repos.

Find me