Open roles
Systems / security engineer
Daemon, hook integration, sandbox orchestration and self-protection. Required: Go, and deep knowledge of Linux or macOS processes, files and networking. Bonus: EDR, proxies, eBPF, container security.
Security research engineer
Threat modeling, red-team corpora, Semgrep rules, the LLM review pipeline and the decoy layer. Required: Python, and offensive or security research experience. Bonus: prompt injection, MCP, agent frameworks.
Full-stack engineer
Console, IDE extensions and approval flows. Required: TypeScript and React, plus a Go or Python backend. Bonus: developer tools.
How we start working together
1 · Send an email
Write to [email protected] with your GitHub, what you've built, the role you want and the hours you can commit each week.
2 · A two-day challenge
In Claude Code, reproduce an agent hijacked by a README injection that reads .env and sends it out, then write the block.
3 · A one-hour video call
Questions in both directions.
4 · A six-week trial
Up to 15 hours a week on one part of the demo: hooks, the action log, decoys or the console.
5 · Sign
When the demo runs, three design partners have installed it and both sides want to continue, we sign the co-founder agreement.

Design partners
Teams can install the demo for free
Engineering teams using AI coding agents can install the demo for free and shape the product. Email [email protected].
Apply
Send your GitHub and the role you want
Write to [email protected] with what you've built, the role you want and the hours you can commit each week. We reply to every message, usually within two days.
