One card per exact version
Bring explicit task terms and submission exports. Earlier and later packets stay separate, even when they belong to the same recipient.
Review the exact submission you received. Keep its evidence, rules and reviewer judgment together, then prepare an ordered allocation preview.
Offline review preview. Not settlement authorization or live eligibility verification.
Bring explicit task terms and submission exports. Earlier and later packets stay separate, even when they belong to the same recipient.
Record dated findings and an explicit reviewer decision. Changes to the reviewed evidence or rules make an old binding inapplicable.
Check recipient caps and integer shares totaling 10,000 basis points. Show unequal remainder shares honestly, with the first chosen recipient marked primary.
The free example uses an entirely fictional promotion task. A reviewed older version stays pinned while a newer incomplete packet remains a separate review.
| Order | Exact submission | Share | State observation |
|---|---|---|---|
| 1 · Primary | demo-A-v1 | 3334 bps · 33.34% | Stale snapshot |
| 2 | demo-B-v1 | 3333 bps · 33.33% | Unknown |
| 3 | demo-C-v1 | 3333 bps · 33.33% | Unknown |
These shares are unequal. No live eligibility or payout is verified.
Qualification is reviewer-authored. A URL does not establish community permission. Missing evidence does not establish misconduct. Imported state can be stale, conflicting or unknown.
The kit provides no wallet, signing payload, acceptance command or settlement action. Local integrity hashes are supporting provenance, not onchain attestation.
67 tests passed on Linux with Python 3.12.14. macOS and other Python versions remain untested. CLI tests do not verify live paid fulfillment.