Developer building small, local-first tools for software workflows.
Interested in AI-assisted coding, code review, dev tooling, maintainable systems, and the gap between “code that works” and “code that should have changed.”
A good harness should not only make agents more capable at completing tasks, but also make their outputs much easier to review.
For example:
A good harness constrains the action surface, context, and task boundaries.
An agent’s failure isn’t always due to “writing incorrect code” — it can also result from “doing things it wasn’t supposed to do.”
Tests and lints can verify part of the correctness, but they often fail to validate task scope.
A well-designed harness should shift the review process from “reading the entire diff” to “verifying whether the changes stay within the defined task boundaries.”
Interested in AI-assisted coding, code review, dev tooling, maintainable systems, and the gap between “code that works” and “code that should have changed.”
GitHub: https://github.com/kestiny18