Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →archive228/sprint-contract
Negotiate a pre-code contract between generator and evaluator personas that defines what "done" means before any code is written. Turns fuzzy specs into a testable target the evaluator can hold the generator to.
Passed 18 of 21 test cases with the skill, and 5 of 21 without it.
and 21 more in the full report
14 fixed5 passed either way3 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 13 test cases the skill fixed. Same model, same prompt, with and without the skill.
Our agent session just rehydrated from `claude-progress.txt` and found an existing contract with 2 unchecked predicate checkboxes. What is the immediate next action according to pre-code sprint guidelines?
| Checked | Without skill | With skill |
|---|---|---|
| The response states to drive the specified runtime path and check off the existing predicates without re-planning. | Missed | Passed |
| The response prohibits modifying or editing the pre-existing contract text. | Missed | Passed |
The instructions your model reads when the skill loads.
The evaluator's leverage collapses when "done" is defined after the code exists. The generator ships something, the evaluator finds it plausible, the fuzzy spec silently reshapes to match what got built. Premature victory, dressed up.
Fix the timing: write the contract before the generator writes a line of code. The evaluator's job for the sprint is then mechanical — hold the artifact against the contract, no re-negotiation mid-flight.
Inspired by the planner/generator/evaluator split in Prithvi's March 2026 post on multi-agent harnesses.
shift-notes / a progress file and needs a concrete acceptance target before it picks up the keyboard.claude-progress.txt (or equivalent) under a ## Sprint contract header, with the sprint start timestamp. Both personas reference this exact text for the rest of the sprint.## Sprint contract — <ISO timestamp>
Deliverable: <one sentence, user-observable>
Acceptance predicates:
- [ ] <predicate 1 — script-decidable>
- [ ] <predicate 2>
- [ ] <predicate 3>
Runtime path: <exact URL / CLI / click sequence the evaluator will drive>
Out of scope this sprint:
- <tempting adjacent fix>
- <tempting refactor>
test cases passed, vs 5 / 21 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.