Tag: agents
RBX Journal articles filed under this tag.
- September 15, 2026
The Bottleneck Moved
AI capability is still advancing. The constraint has moved to system reliability, ground truth and the engineering around the model.
- August 31, 2026
Loop Engineering
58 rounds of adversarial review between two models, with a human on triage, took a runtime migration plan from 12 findings to zero.
- August 23, 2026
When the environment says stop: backpressure as a governance boundary
An agent operating through a browser is a client of a system it does not control. Rate limits and automated-behavior warnings are not obstacles to route around — they are backpressure, and the correct engineering response is to stop, not to accelerate in disguise.
- August 8, 2026
Too much rigor is also a bug
This morning we published the story of the reviewer that rejected its own text. In the afternoon came the next chapter: an entire data exporter built in one day, proven byte by byte against reference vectors, with more than a hundred green tests. And the most valuable finding of the day came from a second reviewer, a twin of the implementer, which found the defect no test in the suite could catch: the code was too strict for the real world.
- August 1, 2026
Governed autonomy is a distributed systems problem
Why idempotency, provenance, unknown outcomes and vendor-neutral observability belong at the center of agentic architecture, not at the margins of the prompt.
- July 7, 2026
Engineering update: the brief a stranger could run
A reconciliation fix led to a second bug nobody asked to find, and the next chapter shipped as a brief written for an agent with zero memory of the day. It ran anyway.
- April 5, 2026
Robson v2 finds its missing execution glue
Over the last three days, Robson v2 gained the governed layer it was missing between risk, approval, audit, projections, and recovery. That is laying real groundwork for v3.
- April 4, 2026
The Unit Is Not the Prompt
Query and QueryEngine point to a more durable abstraction for agent systems: execution is a governed object with state, events, budgets, and permissions.
- April 1, 2026
The Model Is Not the System
A recent AI agent architecture incident reinforced a point most teams still miss: reliable agents are defined by runtime control, not by model selection alone.