Posts tagged "software engineering"

Structured model output is not trusted application input. Validate syntax, semantics, authorization, and effect policy again at the tool execution boundary.

A practical AI failure taxonomy that separates what happened, where it happened, its impact, and the suspected cause so teams can build useful metrics.

A purely random sample hides rare and expensive AI failures. Combine baseline, risk-triggered, and exploratory samples without corrupting the denominator.

A working record-and-replay harness for agent tests: a JSON cassette of decisions and tool results, deterministic replay with zero live calls, and a readable diff when the trajectory changes.

API schema changes are not only field changes. Separate source contracts from product meaning, preserve evidence, and evolve mappings without spreading breakage.