Agents on the desk

The same tooling we build 115 with also helps run it. Optimus, OpenClaw, Hermes, Claude Code, and Codex do research, backtests, and desk operations alongside the work of extending the harness itself. That is the point, and it isn't about novelty: the people building the system and the agents operating it share one loop. Nothing gets thrown over a wall to a separate ops layer. Research feeds tooling, tooling feeds the desk, and the desk feeds back into research, all through the same set of agents.

Day to day, the agents do the bounded, checkable work well. They run backtests and summarize the results. They watch logs and observability, and flag when something drifts. They draft research and check each other's, keep the desk's recurring chores moving, and catch the boring failures early — the stale feed, the job that didn't fire, the number that stopped updating. This is unglamorous work, and it is exactly the work that quietly eats a small desk alive when nobody is watching it.

There is a hard line, and it doesn't move. The agents operate with human-in-the-loop controls: they propose, they prepare, they surface. A person stays on the decisions that move real capital, and on anything that touches risk limits or the live order path. We treat it as a rule, not a preference — an agent that can be wrong should never be the last checkpoint before the market. Agents assemble the context and lay out the options; the person makes the call and takes the action.

For a one-person desk, this is leverage, not autopilot. The agents expand how much ground a single trader can cover: more research explored, more monitored, fewer things falling through the cracks between one session and the next. What they do not do is take over the calls that need judgment. The trader isn't handing off decisions; they are getting more of them surfaced, better prepared, with less noise in the way.

That is the shape we are after — a desk where a human is amplified, not replaced. The agents carry the surface area. The person keeps the wheel, stays close to the risk, and owns every decision that reaches the market. Short feedback loop, clean incentives, and a clear line about who is responsible for what.