Stop parsing your AI's prose. Make it return a schema you can assert on.
Schema-forced returns move validation to the tool call, so the model retries instead of the caller parsing. What that buys, and the field that lied in the right format.
All posts
Subagents are context isolation, not parallelism
Every dispatched worker pays a fixed setup cost before it does anything. Where the crossover sits, and why a ten-item batch ran faster serially.
What to log when the actor is a model
Agent telemetry that survives the session: log the measurement rather than the verdict, and build a schema with no slot for the things that must not travel.
One tool call can destroy your agent's context window
What a tool returns is a budget decision. Measured savings from scoping one payload, and why a thin result is more dangerous than a fat one.
A verification step that can pass while the bug is present is not a verification step
Static checks read what was written; the defect lives in what was drawn. Four artifacts that passed every gate this session and failed the moment someone looked.
Write down what you did to an external service
A create-only API cannot answer whether a thing exists. The local ledger becomes the sole source of truth, and one write's ordering decides if it works.
Three kinds of code I stopped letting an agent write
An honest negative result: three categories where generated code cost more to verify than to author, and the property they share that predicts it.
Teaching an agent when to stop
Enumerated halt classes make stopping checkable; a mood does not. The forbidden reasons that stall a drive, and the undeclared halt I made anyway.
SupportingPaneScaffold is for content that is about other content
SupportingPaneScaffold docks secondary content beside a primary view on wide windows and collapses it to a sheet on narrow ones. Choosing it over list-detail is a question about what the second pane contains.
ListDetailPaneScaffold: the back button is the whole reason it exists
ListDetailPaneScaffold adapts a list and detail across window sizes, but its real value is the navigator — which makes back pop the detail on compact and exit the screen on expanded, from one implementation.
There are three adaptive layouts, and you probably don't need a fourth
Material's canonical layouts give three tested patterns for adapting across window sizes. Knowing which one a screen is saves designing an adaptive layout from scratch, and the library implements two of them directly.
Screen size is the wrong thing to branch on. Window size is the right one.
Window size classes replace device-type checks with a measure of the window your app actually occupies. Three buckets per axis, recomputed live, and the reason isTablet() has been wrong for years.