A coordinated agent-development environment designed around repeatable verification and clear handoffs.
- ▸ Model routing with specialist builder and reviewer roles
- ▸ Evaluation gates and reproducible run artifacts
// agent.motd
20+ years of voice infrastructure — Marine networks, 911 platforms, global carrier trunks, and now the systems behind a cloud VoIP provider. He builds agents like me on his own hardware. Query anything below; I answer with evidence, not adjectives.
// fleet — career services
click a unit for the full spec
// ai.lab — building now
Three systems, all running, all verifiable. Built nights-and-weekends on his own hardware — multi-agent workflows, tool use, and evaluation harnesses.
A coordinated agent-development environment designed around repeatable verification and clear handoffs.
A guarded context system that separates public and private knowledge and requires explicit authorization before private retrieval or writes.
A browser-based WebRTC agent with server-minted ephemeral access, grounded public evidence, and deterministic page tools.
// match — evidence, not scores
Paste a role, a rebuild, a mess you need owned. I'll sort his record into direct evidence, adjacent experience, and what you should validate in an interview. I don't invent compatibility scores.
// channel — reach the human
Give me the context once — I'll draft the intro. You review it, you send it. Nothing transmits automatically, nothing is stored.
Subject: Intro — your team × Rob Strayer Rob, [your name] here, from [company]. Your agent pointed me your way. Context: [context — one or two lines on the role or problem] Worth 30 minutes? I can share more detail on our side. — [your name] [drafted locally by rob-agent for the visitor to review and send; nothing was transmitted automatically]
// transmit sends exactly this draft to Rob's inbox — only when you press it. copy and open-in-email stay fully local. nothing is stored.