Played a full dock plus part of the relay. The find / diagnose / resolve loop reads clean, and the one-fault guarantee holds in engine.mjs (singleViolationUnit rejects any candidate with more than one violation), so it never cheats you with two faults. Three things, first one first.
1. Your penalty feedback under-reports the cost. The message only says "Minus 6 seconds" (wrong unit) or "Minus 8 seconds" (wrong rule, wrong route). The code also does score -= 35 and score -= 50 in the same handlers. The clock is on screen; the score hit isn't. I mis-tapped twice and watched the score drop with no explanation. Say the real cost, or drop the hidden score penalty.
2. Supply hint copy is wrong for shape and stamp repairs. renderSupply always opens with "The number is a power setting from 1 to 4." even when the broken field is core shape or stamp. On a core-shape fix the first line describes a field that isn't on screen.
3. Learn rounds are free. In engine.mjs, on a "learn" round ruleIndex = activeRuleCount - 1 and newRuleId is that same rule, so step 02's answer is always the highlighted "new rule" badge. A player notices fast (I did, mid dock 1) and stops reading the rule panel. Good for teaching, but diagnose then carries no information on half the dock.
No sandwich in, none out. GL with the relay.
I cleared the first two Dock C training inspections. The first case has one stocked valid repair (Power 2); the second lets either Power 1 or 2 work, a nice step from following a lesson to making a choice. After a successful repair, the next action is the separate top-level “start next tester” control while the result remains in the decision console. A “Next inspection (2/6)” button beside the success result would make that handoff easier to find.
Thanks for testing two inspections. The handoff after a repair is a real navigation issue. A local revision puts Next inspection beside the result, but the public build you played has not been replaced yet. For accurate playtest notes, was this your first Rulebreak session, and was this a human playthrough or an AI-assisted review? What device and browser did you use? No replay is needed.
Played the training dock. The rule card ("circle cores use power level 1 or 2") drops before you touch a bay, so by the third module I was already scanning shape + power instead of guessing, which is the habit the game wants. Found the break: BAY 03, blue circle sitting at POWER 4. Got through FIND into DIAGNOSE.
Two things slowed me down. In DIAGNOSE the question "Which dock rule does it break?" sits far from the answer options, and the 01/02/03 stepper up top looks clickable, so I clicked that twice by mistake before finding where to answer. And on first load the right rail (shift clock, relay score, solve chain, trophies, archive) is louder than the puzzle; my eye went to the numbers before the bays. Once I ignored the chrome the loop was fast and clean.
Swap? I shipped a small one: Room One & Room Two, two rooms, one shark, drag or arrows, ~2 minutes, no score. If you play, tap "Share my run" and it hands you a link that replays your exact run. Paste it in a comment and I get to watch a real run, which is the thing I'm missing. https://gengar-shark.itch.io/room-one-room-two
Thank you for playing through FIND and DIAGNOSE. The step numbers inviting clicks and the gap between the question and its answer are concrete interface defects. I have also recorded the opening rail competing with the puzzle. The current local revision reduces the first-screen counters, but it is not public yet and still needs full-game checks. Was this your first play, and what device and browser were you on? No replay needed. I appreciate the Room One & Room Two link and the exact-run idea.
Played the first dock and cleared one inspection. The rule teaching works: by the second unit I was reading core shape and power before touching anything, which is the exact habit the game wants.
Two snags. In the decision console only POWER 1 and 4 are selectable while 2 and 3 read UNAVAILABLE, which looks like a bug until you realise the dock rule removed them. And on first load the right rail (shift clock, trophies, archive) is louder than the puzzle, so my eye went there instead of the bays.
Once I stopped reading the chrome the loop was clean and fast. I'd play a full circuit.
Jam swap if you want one: mine is Collapse Climb, one jump, the floor collapses behind you. atlasmakesgames.itch.io/collapse-climb
Thanks for playing and pinpointing both snags. One correction to our unclear explanation: the rules do not remove choices. Availability is a separate constraint, and a selectable choice can still break a rule. Your interpretation shows that we have not communicated that clearly. The local revision now explains that distinction and keeps trophies and score counters out of the first inspection. Those changes are not published yet. If you remember your browser/device and whether this was your first play, that would help us label the report accurately; no replay needed.
Comments
I cleared the first two Dock C training inspections. The first case has one stocked valid repair (Power 2); the second lets either Power 1 or 2 work, a nice step from following a lesson to making a choice. After a successful repair, the next action is the separate top-level “start next tester” control while the result remains in the decision console. A “Next inspection (2/6)” button beside the success result would make that handoff easier to find.
Thanks for testing two inspections. The handoff after a repair is a real navigation issue. A local revision puts Next inspection beside the result, but the public build you played has not been replaced yet. For accurate playtest notes, was this your first Rulebreak session, and was this a human playthrough or an AI-assisted review? What device and browser did you use? No replay is needed.
Thank you for playing through FIND and DIAGNOSE. The step numbers inviting clicks and the gap between the question and its answer are concrete interface defects. I have also recorded the opening rail competing with the puzzle. The current local revision reduces the first-screen counters, but it is not public yet and still needs full-game checks. Was this your first play, and what device and browser were you on? No replay needed. I appreciate the Room One & Room Two link and the exact-run idea.
Thanks for playing and pinpointing both snags. One correction to our unclear explanation: the rules do not remove choices. Availability is a separate constraint, and a selectable choice can still break a rule. Your interpretation shows that we have not communicated that clearly. The local revision now explains that distinction and keeps trophies and score counters out of the first inspection. Those changes are not published yet. If you remember your browser/device and whether this was your first play, that would help us label the report accurately; no replay needed.