Opinionated harness per work type

The idea

Each work template carries steps, a quality bar, pitfalls, tool guidance and practitioner notes, plus a minimum model tier and effort. Its named industry checklists are settled by detectors or by attestations with evidence.

The problem
A generic worker does not know the practice or the stakes of each kind of job.
Why it matters
Each built-in work type starts with a stated bar, its checks and a model floor.
Where it stands
In production as of

A step-by-step replay of this idea is not published yet. The library adds one when a recorded run shows it.

Where it lives in the product

  • Self-improving template harness

    After work finishes, two proposers edit the template's harness from the run ledger. Each candidate is probed on the same request. A composer and a blind selector pick one, which a person adopts (or auto-adopt does, if the checks pass).

  • Tier × effort routing from runtime signals

    Each call's model tier and effort are chosen separately. The inputs are the stage, the risk of the pending action, the change size, open findings, failures, repository size and the budget left. Floors tune from verified outcomes.

  • Harness tools that run checks and keep evidence

    QA test cases run in a headless browser on the item's own previews, with screenshots. Runbooks run as drills with approval and rollback. Device flashes are checked by a read-back hash.

  • Acceptance criteria as executable checks

    At plan time, each acceptance criterion maps to a test, script, query, HTTP check or screenshot check. Verification runs all of them in each round.

Turn the numbers into shipped work.

Crewshal runs these choices for you: a persistent AI worker with memory and rules, on your Claude and Codex subscriptions.