Summary
Nate B Jones walks through Open Engine, a personal orchestration layer he built to coordinate Claude, Codex, ChatGPT, and OpenClaw (Hermes) without waiting for official integrations between these tools. The core problem he addresses is the “handoff problem”: work created in one AI system rarely transfers cleanly to another, forcing users to serve as manual context carriers across five or more subscriptions. Jones argues this coordination overhead is the primary bottleneck limiting AI productivity in 2026, not model capability or user fluency.
Open Engine uses Linear as a task-tracking backbone (with Jira as a drop-in alternative) paired with four skill files that teach each AI how to interact with the ticketing system: a setup skill, a status skill, a queue-running skill, and a smoke test. The architecture is designed to let work leave Claude and land with Codex, allow a teammate’s agent to pick up tasks created by another agent, and let support escalations reach the right human without losing message history or customer context. Jones emphasizes that Open Engine is not anti-OpenClaw or anti-Hermes — it is designed to work alongside them.
The video includes concrete personal use cases: managing a house move, publishing stories, and helping a friend who runs an agency while raising a baby coordinate client calls, product scoping sessions, and calendar juggling across Claude Code and Codex. Jones is releasing Open Engine as an open-source project, positioning it for both solo power users and small teams where human and AI agents need to hand off work fluidly.
📺 Source: AI News & Strategy Daily | Nate B Jones · Published June 26, 2026
🏷️ Format: Hands On Build
