Mechanics design review record

The earlier review below covers the site-contest alternative and its proposed set. That alternative is not the selected browser prototype. The current duel received a separate review, recorded next.

Current duel review

On 3 October 2026, Claude Opus 5.5 reviewed the compact duel rules with tools disabled. It explicitly concluded: “Yes, this is acceptable for a first playable test”, finding no contradiction that blocked the prototype. The review covered the written core rules, not the actual card file, implementation, or human play.

Clarifications now explicitly state that attacks exhaust units, Guard restricts attacks only, a targetless Arrival does not prevent deployment, and hero defeat is checked after full action resolution or turn bookkeeping. The starting pool excludes health modifiers, generated units, and extra timing triggers. The reviewer flagged first-player advantage, board snowballing, defensive stalling, and a familiar identity as playtest questions. No added subsystem is prescribed to solve those questions before play.

Earlier site-contest review

Review date: 3 October 2026. Scope: mechanics candidates and pilot rules and the initial set structure and starter cards. The review checks whether the written pilot can be played as a tabletop game and whether the proposed set arithmetic is coherent. It does not establish enjoyable play, match duration, strategic diversity, or balance.

The review used Claude Opus 5.5 through claude -p --model claude-opus-5-5 --tools '' --no-session-persistence, with the full documents and the user's brief supplied as input. Tools were disabled. The reviewer was asked to identify real defects and give an independent acceptance or rejection, distinguishing rules gaps from questions requiring playtests.

First review and revisions

The first verdict was “Acceptable for the first playtest, after four small changes.” This was conditional acceptance, so the documents were revised and returned for another review.

The reviewer independently checked the 100-card partitions, 21 distinct pilot designs, two 30-card decks, type counts, 36 action opportunities, and the maximum 36 points from sites. Local arithmetic checks agreed.

The review also exposed a property to observe in play: round one allows at most one useful action per player, so the other opportunities become passes. Those passes may be strategically chosen before deployment or forced afterward. This remains explicitly documented in the baseline. A setup variant is proposed only if testing shows the opening feels wasted. Whether banking, persistent damage, relics, or last-action boosts are enjoyable remains a playtest question.

Second review and revisions

The second review explicitly accepted the pilot for its first tabletop playtest. It also identified further improvements to the provisional set plan. The remaining neutral six-cost slot moved to cost four, leaving one six-cost Growth card. Returning a unit to hand is now explicitly distinct from repeating an Arrival for free: replay requires a normal paid action and legal space. This makes future Mobility text compatible with the loop restrictions.

The teaching outline now includes the scoring schedule, tied sites, defeat threshold, initiative, relic replacement, and opening redraw. The optional setup variant specifies timing, hidden simultaneous choices, and the supply cap. Protection's comparison against the pilot defenses and the already-filled neutral two-cost budget are recorded as requirements for the next development pool.

Four temporary pilot names were changed to Vanguard Recruit, Field Medic, Heavy Vanguard, and Forward Standard to avoid treating an unchosen theme as settled. Their rules did not change. These edits were included in the final review input.

Final status

The third review explicitly concluded: “The first-playtest rules are accepted” and “The 100-card plan is accepted as a provisional plan.” This ends the requested design-review loop; acceptance is not evidence of playtest success.

The reviewer found one test-procedure defect: swapping both players' decks and starting player in each two-game pair would leave the same deck starting twice. The recorded-game plan now includes a fixed six-game table, overrides random opening initiative for those games, and gives each deck three starts. Its remaining four-to-two starting-player split is disclosed and reversed in the next batch. The table was checked directly.

Additional reference clarifications from that final review specify that supply is permanent, leaving play clears damage and effects, a tactic's temporary bonus survives its own discard, and the optional setup variant advances all card costs. The set plan's late-cost statements now explicitly refer to the baseline. These were clarifications after the accepted design review, not a claim of another review round.

No human playtest has been performed. Fun, duration, learnability, starting-player advantage, strategy breadth, and balance remain unverified. The next evidence should come from playing the small pilot, rather than another paper review or a larger card file.