Thinking on software identity, agentic development, and the engineering decisions that persist.
In 2025, the hard problem was finding enough context for agents. In 2026, the hard problem is finding less of it. When identity exists upstream, the new skill is slicing -- carving the smallest coherent cross-section of DNA that contains all the concerns for a given task.
The biggest fear CTOs have about identity engineering is that it requires pausing feature work. It doesn't. The strangler fig pattern lets you start with one service, prove the value in a single quarter, and expand without stopping the roadmap.
Identity engineering has real failure modes. Teams that adopt it badly end up worse than teams that never adopted it at all. This article names the three most common ways it goes wrong and the antidote for each.
Identity gaps have a dollar cost. Not a vague 'technical debt' cost -- an actual, calculable cost per quarter that shows up in onboarding delays, incident duration, review bottlenecks, agent rejection rates, and compliance overhead. Here is the model to calculate it for your own org.
Teams that declared what their software is supposed to be, then measured against it, keep reporting the same pattern: onboarding cuts in half, production outages drop to near zero, deploy frequency doubles. These are the closest real-world numbers we have for identity engineering, and they tell a consistent story.
Identity engineering has to fit into the workflow you already have or it will be rejected. This article maps exactly where identity declarations get created, reviewed, and updated within the sprint cycle, PR workflow, CI/CD pipeline, and incident response.
Most engineering teams already have fragments of an identity layer scattered across Backstage, OpenAPI, Terraform, CLAUDE.md, ADRs, and K8s manifests. The problem isn't that you have no identity layer. It's that you have half of one, fragmented across tools that don't talk to each other, with critical gaps between them.
Your immune system eliminates almost every T-cell it produces -- not for being broken, but for failing an identity check. Software teams pentest behavior but almost never pentest identity. That gap is where the hardest bugs live.
45% of software defects originate before anyone writes a line of code. You can't catch them with better tests because the tests inherit the same blind spots as the spec. The fix is testing upstream: probing what the software claims to be, not just what it does.
Biological systems converge on optimal solutions because their targets have shape. Slogans don't converge. Specific, measurable identity declarations do. The difference determines whether your agent fleet finds the architecture or wanders forever.
We use cookies to understand how you use ribo.dev and improve your experience.
Learn more in our Cookie Policy