Spec-driven development
Writing the specification and acceptance tests first, then letting agents implement against them.
The human effort moves from typing code to deciding what correct means. A good spec names inputs, outputs, edge cases and how success is checked. Agents then iterate until the checks pass, and a reviewer reads a smaller, clearer diff.
Related terms
- Coding agent: An agent that reads a repository, edits files, runs commands and tests, and opens pull requests.
- Vibe coding: Building by prompting and accepting the result without reading the code closely.
- Evals: A repeatable test set that scores an AI feature, so changes can be measured instead of guessed.
