@opengamehq/skills 1.4.0 → 1.5.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (382) hide show
  1. package/CHANGELOG.md +29 -0
  2. package/README.md +12 -4
  3. package/dist/.claude-plugin/marketplace.json +26 -0
  4. package/dist/skills/manifest.json +1406 -142
  5. package/dist/skills/opengame-accessibility/CHANGELOG.md +4 -0
  6. package/dist/skills/opengame-accessibility/SKILL.md +48 -32
  7. package/dist/skills/opengame-all-in-one/CHANGELOG.md +21 -0
  8. package/dist/skills/opengame-all-in-one/SKILL.md +195 -0
  9. package/dist/skills/opengame-all-in-one/references/opengame-accessibility/README.md +102 -0
  10. package/dist/skills/opengame-all-in-one/references/opengame-api/README.md +107 -0
  11. package/dist/skills/opengame-all-in-one/references/opengame-assets/README.md +168 -0
  12. package/dist/skills/opengame-all-in-one/references/opengame-assets/references/pipeline.md +164 -0
  13. package/dist/skills/opengame-all-in-one/references/opengame-audio/README.md +238 -0
  14. package/dist/skills/opengame-all-in-one/references/opengame-audio/helpers/audio-chain.js +325 -0
  15. package/dist/skills/opengame-all-in-one/references/opengame-audio/references/mix-notes.md +125 -0
  16. package/dist/skills/opengame-all-in-one/references/opengame-audio/references/qa-without-ears.md +177 -0
  17. package/dist/skills/opengame-all-in-one/references/opengame-capture/README.md +128 -0
  18. package/dist/skills/opengame-all-in-one/references/opengame-capture/references/pipeline.md +114 -0
  19. package/dist/skills/opengame-all-in-one/references/opengame-capture/references/threejs.md +161 -0
  20. package/dist/skills/opengame-all-in-one/references/opengame-characters/README.md +180 -0
  21. package/dist/skills/opengame-all-in-one/references/opengame-characters/references/first-person-body.md +108 -0
  22. package/dist/skills/opengame-all-in-one/references/opengame-characters/references/integration.md +144 -0
  23. package/dist/skills/opengame-all-in-one/references/opengame-characters/references/player-og.md +160 -0
  24. package/dist/skills/opengame-all-in-one/references/opengame-cli/README.md +189 -0
  25. package/dist/skills/opengame-all-in-one/references/opengame-deploy/README.md +223 -0
  26. package/dist/skills/opengame-all-in-one/references/opengame-deploy/references/json-events.md +63 -0
  27. package/dist/skills/opengame-all-in-one/references/opengame-deploy/references/manifest.md +89 -0
  28. package/dist/skills/opengame-all-in-one/references/opengame-deploy/references/packaging.md +119 -0
  29. package/dist/skills/opengame-all-in-one/references/opengame-depth/README.md +154 -0
  30. package/dist/skills/opengame-all-in-one/references/opengame-enemies/README.md +155 -0
  31. package/dist/skills/opengame-all-in-one/references/opengame-enemies/references/pacing-and-bots.md +108 -0
  32. package/dist/skills/opengame-all-in-one/references/opengame-enemies/references/threejs.md +193 -0
  33. package/dist/skills/opengame-all-in-one/references/opengame-front-end/README.md +185 -0
  34. package/dist/skills/opengame-all-in-one/references/opengame-front-end/references/launch.md +50 -0
  35. package/dist/skills/opengame-all-in-one/references/opengame-front-end/references/menus.md +81 -0
  36. package/dist/skills/opengame-all-in-one/references/opengame-game-feel/README.md +181 -0
  37. package/dist/skills/opengame-all-in-one/references/opengame-game-feel/helpers/spring.js +146 -0
  38. package/dist/skills/opengame-all-in-one/references/opengame-game-feel/references/feel-numbers.md +142 -0
  39. package/dist/skills/opengame-all-in-one/references/opengame-game-feel/references/killcam.md +88 -0
  40. package/dist/skills/opengame-all-in-one/references/opengame-jam/README.md +114 -0
  41. package/dist/skills/opengame-all-in-one/references/opengame-lighting/README.md +141 -0
  42. package/dist/skills/opengame-all-in-one/references/opengame-lighting/helpers/light-pool.js +185 -0
  43. package/dist/skills/opengame-all-in-one/references/opengame-loading/README.md +135 -0
  44. package/dist/skills/opengame-all-in-one/references/opengame-loading/helpers/loading-count.mjs +115 -0
  45. package/dist/skills/opengame-all-in-one/references/opengame-loading/references/threejs.md +141 -0
  46. package/dist/skills/opengame-all-in-one/references/opengame-look/README.md +181 -0
  47. package/dist/skills/opengame-all-in-one/references/opengame-look/references/matching-a-reference.md +73 -0
  48. package/dist/skills/opengame-all-in-one/references/opengame-look/references/painterly.md +69 -0
  49. package/dist/skills/opengame-all-in-one/references/opengame-look/references/post-and-lens.md +73 -0
  50. package/dist/skills/opengame-all-in-one/references/opengame-maps/README.md +74 -0
  51. package/dist/skills/opengame-all-in-one/references/opengame-maps/references/map-spec.md +77 -0
  52. package/dist/skills/opengame-all-in-one/references/opengame-mobile/README.md +86 -0
  53. package/dist/skills/opengame-all-in-one/references/opengame-montage-make/README.md +183 -0
  54. package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/briefs.md +293 -0
  55. package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/day-one-questions.md +46 -0
  56. package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/edge-cases.md +27 -0
  57. package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/filmable.md +244 -0
  58. package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/finishing.md +126 -0
  59. package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/measure.md +282 -0
  60. package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/montage-capture.md +151 -0
  61. package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/montage-gates.md +84 -0
  62. package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/montage-genres.md +198 -0
  63. package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/montage-grammar.md +320 -0
  64. package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/montage-ops.md +127 -0
  65. package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/og-snipers-record.md +110 -0
  66. package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/owner-eye.md +214 -0
  67. package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/reference-study.md +293 -0
  68. package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/song.md +64 -0
  69. package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/README.md +193 -0
  70. package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/briefs.md +293 -0
  71. package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/day-one-questions.md +46 -0
  72. package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/filmable.md +244 -0
  73. package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/measure.md +282 -0
  74. package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/montage-capture.md +151 -0
  75. package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/montage-gates.md +84 -0
  76. package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/montage-genres.md +198 -0
  77. package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/montage-grammar.md +320 -0
  78. package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/montage-ops.md +127 -0
  79. package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/og-snipers-record.md +110 -0
  80. package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/owner-eye.md +214 -0
  81. package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/reference-study.md +293 -0
  82. package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/song.md +64 -0
  83. package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/README.md +174 -0
  84. package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/briefs.md +293 -0
  85. package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/day-one-questions.md +46 -0
  86. package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/filmable.md +244 -0
  87. package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/finishing.md +126 -0
  88. package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/measure.md +282 -0
  89. package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/montage-capture.md +151 -0
  90. package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/montage-gates.md +84 -0
  91. package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/montage-genres.md +198 -0
  92. package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/montage-grammar.md +320 -0
  93. package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/montage-ops.md +127 -0
  94. package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/og-snipers-record.md +110 -0
  95. package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/owner-eye.md +214 -0
  96. package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/reference-study.md +293 -0
  97. package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/song.md +64 -0
  98. package/dist/skills/opengame-all-in-one/references/opengame-multiplayer/README.md +139 -0
  99. package/dist/skills/opengame-all-in-one/references/opengame-multiplayer/references/bot-lobbies.md +71 -0
  100. package/dist/skills/opengame-all-in-one/references/opengame-multiplayer/references/netcode.md +165 -0
  101. package/dist/skills/opengame-all-in-one/references/opengame-performance/README.md +177 -0
  102. package/dist/skills/opengame-all-in-one/references/opengame-performance/helpers/landing-pacer.mjs +86 -0
  103. package/dist/skills/opengame-all-in-one/references/opengame-performance/references/measuring.md +123 -0
  104. package/dist/skills/opengame-all-in-one/references/opengame-performance/references/webgpu.md +86 -0
  105. package/dist/skills/opengame-all-in-one/references/opengame-production/README.md +192 -0
  106. package/dist/skills/opengame-all-in-one/references/opengame-production/references/team.md +107 -0
  107. package/dist/skills/opengame-all-in-one/references/opengame-production/references/templates.md +153 -0
  108. package/dist/skills/opengame-all-in-one/references/opengame-quality-bar/README.md +170 -0
  109. package/dist/skills/opengame-all-in-one/references/opengame-quality-bar/references/cheap-tells.md +43 -0
  110. package/dist/skills/opengame-all-in-one/references/opengame-quality-bar/references/judges.md +95 -0
  111. package/dist/skills/opengame-all-in-one/references/opengame-quality-bar/references/review-prompt.md +32 -0
  112. package/dist/skills/opengame-all-in-one/references/opengame-rhythm/README.md +134 -0
  113. package/dist/skills/opengame-all-in-one/references/opengame-rhythm/references/charting.md +109 -0
  114. package/dist/skills/opengame-all-in-one/references/opengame-rhythm/references/timing.md +92 -0
  115. package/dist/skills/opengame-all-in-one/references/opengame-router/README.md +54 -0
  116. package/dist/skills/opengame-all-in-one/references/opengame-router/references/process.md +38 -0
  117. package/dist/skills/opengame-all-in-one/references/opengame-scene-audit/README.md +136 -0
  118. package/dist/skills/opengame-all-in-one/references/opengame-scene-audit/helpers/scene-audit.js +677 -0
  119. package/dist/skills/opengame-all-in-one/references/opengame-sdk/README.md +283 -0
  120. package/dist/skills/opengame-all-in-one/references/opengame-sdk/references/engines.md +93 -0
  121. package/dist/skills/opengame-all-in-one/references/opengame-sdk/references/frame.md +107 -0
  122. package/dist/skills/opengame-all-in-one/references/opengame-sdk/references/identity.md +111 -0
  123. package/dist/skills/opengame-all-in-one/references/opengame-sdk/references/leaderboards.md +115 -0
  124. package/dist/skills/opengame-all-in-one/references/opengame-sdk/references/radio.md +71 -0
  125. package/dist/skills/opengame-all-in-one/references/opengame-stability/README.md +112 -0
  126. package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/README.md +160 -0
  127. package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/archetypes.md +113 -0
  128. package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/briefs.md +273 -0
  129. package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/core.md +129 -0
  130. package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/craft.md +299 -0
  131. package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/edge-cases.md +27 -0
  132. package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/filmable.md +244 -0
  133. package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/finishing.md +126 -0
  134. package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/gates.md +184 -0
  135. package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/measure.md +282 -0
  136. package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/music.md +148 -0
  137. package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/og-clash-record.md +133 -0
  138. package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/ops.md +154 -0
  139. package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/owner-eye.md +214 -0
  140. package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/review.md +141 -0
  141. package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/title.md +92 -0
  142. package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/README.md +172 -0
  143. package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/archetypes.md +113 -0
  144. package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/briefs.md +273 -0
  145. package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/core.md +129 -0
  146. package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/craft.md +299 -0
  147. package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/edge-cases.md +27 -0
  148. package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/filmable.md +244 -0
  149. package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/finishing.md +126 -0
  150. package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/gates.md +184 -0
  151. package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/measure.md +282 -0
  152. package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/music.md +148 -0
  153. package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/og-clash-record.md +133 -0
  154. package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/ops.md +154 -0
  155. package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/owner-eye.md +214 -0
  156. package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/reference-set.md +92 -0
  157. package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/review.md +141 -0
  158. package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/title.md +92 -0
  159. package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/README.md +141 -0
  160. package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/archetypes.md +113 -0
  161. package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/briefs.md +273 -0
  162. package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/core.md +129 -0
  163. package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/craft.md +299 -0
  164. package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/edge-cases.md +27 -0
  165. package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/filmable.md +244 -0
  166. package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/finishing.md +126 -0
  167. package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/gates.md +184 -0
  168. package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/measure.md +282 -0
  169. package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/music.md +148 -0
  170. package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/og-clash-record.md +133 -0
  171. package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/ops.md +154 -0
  172. package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/owner-eye.md +214 -0
  173. package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/review.md +141 -0
  174. package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/title.md +92 -0
  175. package/dist/skills/opengame-all-in-one/references/opengame-verification/README.md +197 -0
  176. package/dist/skills/opengame-all-in-one/references/opengame-verification/helpers/browser-evidence.mjs +74 -0
  177. package/dist/skills/opengame-all-in-one/references/opengame-weapons/README.md +179 -0
  178. package/dist/skills/opengame-all-in-one/references/opengame-weapons/references/handling-numbers.md +114 -0
  179. package/dist/skills/opengame-all-in-one/references/opengame-weapons/references/reloads-and-hands.md +96 -0
  180. package/dist/skills/opengame-all-in-one/references/opengame-weapons/references/sights-and-scopes.md +98 -0
  181. package/dist/skills/opengame-all-in-one/references/opengame-weapons/references/threejs.md +240 -0
  182. package/dist/skills/opengame-all-in-one/references/opengame-world-physics/README.md +141 -0
  183. package/dist/skills/opengame-all-in-one/references/opengame-world-physics/references/measured-failures.md +79 -0
  184. package/dist/skills/opengame-api/CHANGELOG.md +8 -0
  185. package/dist/skills/opengame-api/SKILL.md +39 -13
  186. package/dist/skills/opengame-assets/CHANGELOG.md +11 -0
  187. package/dist/skills/opengame-assets/SKILL.md +79 -31
  188. package/dist/skills/opengame-assets/references/pipeline.md +67 -0
  189. package/dist/skills/opengame-audio/CHANGELOG.md +11 -0
  190. package/dist/skills/opengame-audio/SKILL.md +129 -41
  191. package/dist/skills/opengame-audio/references/mix-notes.md +44 -1
  192. package/dist/skills/opengame-audio/references/qa-without-ears.md +177 -0
  193. package/dist/skills/opengame-capture/CHANGELOG.md +10 -0
  194. package/dist/skills/opengame-capture/SKILL.md +44 -20
  195. package/dist/skills/opengame-capture/references/pipeline.md +33 -1
  196. package/dist/skills/opengame-capture/references/threejs.md +17 -1
  197. package/dist/skills/opengame-characters/CHANGELOG.md +15 -0
  198. package/dist/skills/opengame-characters/SKILL.md +94 -39
  199. package/dist/skills/opengame-characters/references/first-person-body.md +108 -0
  200. package/dist/skills/opengame-characters/references/player-og.md +1 -1
  201. package/dist/skills/opengame-cli/CHANGELOG.md +12 -0
  202. package/dist/skills/opengame-cli/SKILL.md +11 -10
  203. package/dist/skills/opengame-deploy/CHANGELOG.md +20 -0
  204. package/dist/skills/opengame-deploy/SKILL.md +41 -21
  205. package/dist/skills/opengame-deploy/references/json-events.md +6 -5
  206. package/dist/skills/opengame-deploy/references/packaging.md +28 -26
  207. package/dist/skills/opengame-depth/CHANGELOG.md +8 -0
  208. package/dist/skills/opengame-depth/SKILL.md +61 -33
  209. package/dist/skills/opengame-enemies/CHANGELOG.md +10 -0
  210. package/dist/skills/opengame-enemies/SKILL.md +96 -40
  211. package/dist/skills/opengame-enemies/references/pacing-and-bots.md +108 -0
  212. package/dist/skills/opengame-enemies/references/threejs.md +59 -0
  213. package/dist/skills/opengame-front-end/CHANGELOG.md +4 -0
  214. package/dist/skills/opengame-front-end/SKILL.md +85 -39
  215. package/dist/skills/opengame-front-end/references/menus.md +42 -8
  216. package/dist/skills/opengame-game-feel/CHANGELOG.md +11 -0
  217. package/dist/skills/opengame-game-feel/SKILL.md +95 -47
  218. package/dist/skills/opengame-game-feel/references/feel-numbers.md +59 -0
  219. package/dist/skills/opengame-game-feel/references/killcam.md +88 -0
  220. package/dist/skills/opengame-jam/CHANGELOG.md +8 -0
  221. package/dist/skills/opengame-jam/SKILL.md +46 -31
  222. package/dist/skills/opengame-lighting/CHANGELOG.md +9 -0
  223. package/dist/skills/opengame-lighting/SKILL.md +85 -34
  224. package/dist/skills/opengame-loading/CHANGELOG.md +4 -0
  225. package/dist/skills/opengame-loading/SKILL.md +78 -38
  226. package/dist/skills/opengame-loading/helpers/loading-count.mjs +115 -0
  227. package/dist/skills/opengame-loading/references/threejs.md +32 -7
  228. package/dist/skills/opengame-look/CHANGELOG.md +9 -0
  229. package/dist/skills/opengame-look/SKILL.md +122 -51
  230. package/dist/skills/opengame-look/references/matching-a-reference.md +73 -0
  231. package/dist/skills/opengame-look/references/painterly.md +69 -0
  232. package/dist/skills/opengame-look/references/post-and-lens.md +10 -3
  233. package/dist/skills/opengame-maps/CHANGELOG.md +9 -0
  234. package/dist/skills/opengame-maps/SKILL.md +78 -0
  235. package/dist/skills/opengame-maps/references/map-spec.md +77 -0
  236. package/dist/skills/opengame-mobile/CHANGELOG.md +4 -0
  237. package/dist/skills/opengame-mobile/SKILL.md +18 -18
  238. package/dist/skills/opengame-montage-make/CHANGELOG.md +5 -0
  239. package/dist/skills/opengame-montage-make/SKILL.md +187 -0
  240. package/dist/skills/opengame-montage-make/references/briefs.md +293 -0
  241. package/dist/skills/opengame-montage-make/references/day-one-questions.md +46 -0
  242. package/dist/skills/opengame-montage-make/references/edge-cases.md +27 -0
  243. package/dist/skills/opengame-montage-make/references/filmable.md +244 -0
  244. package/dist/skills/opengame-montage-make/references/finishing.md +126 -0
  245. package/dist/skills/opengame-montage-make/references/measure.md +282 -0
  246. package/dist/skills/opengame-montage-make/references/montage-capture.md +151 -0
  247. package/dist/skills/opengame-montage-make/references/montage-gates.md +84 -0
  248. package/dist/skills/opengame-montage-make/references/montage-genres.md +198 -0
  249. package/dist/skills/opengame-montage-make/references/montage-grammar.md +320 -0
  250. package/dist/skills/opengame-montage-make/references/montage-ops.md +127 -0
  251. package/dist/skills/opengame-montage-make/references/og-snipers-record.md +110 -0
  252. package/dist/skills/opengame-montage-make/references/owner-eye.md +214 -0
  253. package/dist/skills/opengame-montage-make/references/reference-study.md +293 -0
  254. package/dist/skills/opengame-montage-make/references/song.md +64 -0
  255. package/dist/skills/opengame-montage-plan/CHANGELOG.md +5 -0
  256. package/dist/skills/opengame-montage-plan/SKILL.md +197 -0
  257. package/dist/skills/opengame-montage-plan/references/briefs.md +293 -0
  258. package/dist/skills/opengame-montage-plan/references/day-one-questions.md +46 -0
  259. package/dist/skills/opengame-montage-plan/references/filmable.md +244 -0
  260. package/dist/skills/opengame-montage-plan/references/measure.md +282 -0
  261. package/dist/skills/opengame-montage-plan/references/montage-capture.md +151 -0
  262. package/dist/skills/opengame-montage-plan/references/montage-gates.md +84 -0
  263. package/dist/skills/opengame-montage-plan/references/montage-genres.md +198 -0
  264. package/dist/skills/opengame-montage-plan/references/montage-grammar.md +320 -0
  265. package/dist/skills/opengame-montage-plan/references/montage-ops.md +127 -0
  266. package/dist/skills/opengame-montage-plan/references/og-snipers-record.md +110 -0
  267. package/dist/skills/opengame-montage-plan/references/owner-eye.md +214 -0
  268. package/dist/skills/opengame-montage-plan/references/reference-study.md +293 -0
  269. package/dist/skills/opengame-montage-plan/references/song.md +64 -0
  270. package/dist/skills/opengame-montage-review-and-improve/CHANGELOG.md +5 -0
  271. package/dist/skills/opengame-montage-review-and-improve/SKILL.md +178 -0
  272. package/dist/skills/opengame-montage-review-and-improve/references/briefs.md +293 -0
  273. package/dist/skills/opengame-montage-review-and-improve/references/day-one-questions.md +46 -0
  274. package/dist/skills/opengame-montage-review-and-improve/references/filmable.md +244 -0
  275. package/dist/skills/opengame-montage-review-and-improve/references/finishing.md +126 -0
  276. package/dist/skills/opengame-montage-review-and-improve/references/measure.md +282 -0
  277. package/dist/skills/opengame-montage-review-and-improve/references/montage-capture.md +151 -0
  278. package/dist/skills/opengame-montage-review-and-improve/references/montage-gates.md +84 -0
  279. package/dist/skills/opengame-montage-review-and-improve/references/montage-genres.md +198 -0
  280. package/dist/skills/opengame-montage-review-and-improve/references/montage-grammar.md +320 -0
  281. package/dist/skills/opengame-montage-review-and-improve/references/montage-ops.md +127 -0
  282. package/dist/skills/opengame-montage-review-and-improve/references/og-snipers-record.md +110 -0
  283. package/dist/skills/opengame-montage-review-and-improve/references/owner-eye.md +214 -0
  284. package/dist/skills/opengame-montage-review-and-improve/references/reference-study.md +293 -0
  285. package/dist/skills/opengame-montage-review-and-improve/references/song.md +64 -0
  286. package/dist/skills/opengame-multiplayer/CHANGELOG.md +15 -0
  287. package/dist/skills/opengame-multiplayer/SKILL.md +143 -0
  288. package/dist/skills/opengame-multiplayer/references/bot-lobbies.md +71 -0
  289. package/dist/skills/opengame-multiplayer/references/netcode.md +165 -0
  290. package/dist/skills/opengame-performance/CHANGELOG.md +8 -0
  291. package/dist/skills/opengame-performance/SKILL.md +106 -65
  292. package/dist/skills/opengame-performance/helpers/landing-pacer.mjs +86 -0
  293. package/dist/skills/opengame-performance/references/measuring.md +51 -0
  294. package/dist/skills/opengame-performance/references/webgpu.md +86 -0
  295. package/dist/skills/opengame-production/CHANGELOG.md +13 -0
  296. package/dist/skills/opengame-production/SKILL.md +119 -54
  297. package/dist/skills/opengame-production/references/team.md +107 -0
  298. package/dist/skills/opengame-production/references/templates.md +106 -16
  299. package/dist/skills/opengame-quality-bar/CHANGELOG.md +10 -0
  300. package/dist/skills/opengame-quality-bar/SKILL.md +94 -42
  301. package/dist/skills/opengame-quality-bar/references/cheap-tells.md +14 -7
  302. package/dist/skills/opengame-quality-bar/references/judges.md +95 -0
  303. package/dist/skills/opengame-quality-bar/references/review-prompt.md +10 -5
  304. package/dist/skills/opengame-rhythm/CHANGELOG.md +5 -0
  305. package/dist/skills/opengame-rhythm/SKILL.md +138 -0
  306. package/dist/skills/opengame-rhythm/references/charting.md +109 -0
  307. package/dist/skills/opengame-rhythm/references/timing.md +92 -0
  308. package/dist/skills/opengame-router/CHANGELOG.md +8 -0
  309. package/dist/skills/opengame-router/SKILL.md +14 -7
  310. package/dist/skills/opengame-scene-audit/CHANGELOG.md +15 -0
  311. package/dist/skills/opengame-scene-audit/SKILL.md +64 -47
  312. package/dist/skills/opengame-scene-audit/helpers/scene-audit.js +97 -19
  313. package/dist/skills/opengame-sdk/CHANGELOG.md +12 -0
  314. package/dist/skills/opengame-sdk/SKILL.md +26 -9
  315. package/dist/skills/opengame-sdk/references/engines.md +2 -2
  316. package/dist/skills/opengame-sdk/references/identity.md +10 -11
  317. package/dist/skills/opengame-stability/CHANGELOG.md +4 -0
  318. package/dist/skills/opengame-stability/SKILL.md +59 -27
  319. package/dist/skills/opengame-trailer-make/CHANGELOG.md +5 -0
  320. package/dist/skills/opengame-trailer-make/SKILL.md +164 -0
  321. package/dist/skills/opengame-trailer-make/references/archetypes.md +113 -0
  322. package/dist/skills/opengame-trailer-make/references/briefs.md +273 -0
  323. package/dist/skills/opengame-trailer-make/references/core.md +129 -0
  324. package/dist/skills/opengame-trailer-make/references/craft.md +299 -0
  325. package/dist/skills/opengame-trailer-make/references/edge-cases.md +27 -0
  326. package/dist/skills/opengame-trailer-make/references/filmable.md +244 -0
  327. package/dist/skills/opengame-trailer-make/references/finishing.md +126 -0
  328. package/dist/skills/opengame-trailer-make/references/gates.md +184 -0
  329. package/dist/skills/opengame-trailer-make/references/measure.md +282 -0
  330. package/dist/skills/opengame-trailer-make/references/music.md +148 -0
  331. package/dist/skills/opengame-trailer-make/references/og-clash-record.md +133 -0
  332. package/dist/skills/opengame-trailer-make/references/ops.md +154 -0
  333. package/dist/skills/opengame-trailer-make/references/owner-eye.md +214 -0
  334. package/dist/skills/opengame-trailer-make/references/review.md +141 -0
  335. package/dist/skills/opengame-trailer-make/references/title.md +92 -0
  336. package/dist/skills/opengame-trailer-plan/CHANGELOG.md +5 -0
  337. package/dist/skills/opengame-trailer-plan/SKILL.md +176 -0
  338. package/dist/skills/opengame-trailer-plan/references/archetypes.md +113 -0
  339. package/dist/skills/opengame-trailer-plan/references/briefs.md +273 -0
  340. package/dist/skills/opengame-trailer-plan/references/core.md +129 -0
  341. package/dist/skills/opengame-trailer-plan/references/craft.md +299 -0
  342. package/dist/skills/opengame-trailer-plan/references/edge-cases.md +27 -0
  343. package/dist/skills/opengame-trailer-plan/references/filmable.md +244 -0
  344. package/dist/skills/opengame-trailer-plan/references/finishing.md +126 -0
  345. package/dist/skills/opengame-trailer-plan/references/gates.md +184 -0
  346. package/dist/skills/opengame-trailer-plan/references/measure.md +282 -0
  347. package/dist/skills/opengame-trailer-plan/references/music.md +148 -0
  348. package/dist/skills/opengame-trailer-plan/references/og-clash-record.md +133 -0
  349. package/dist/skills/opengame-trailer-plan/references/ops.md +154 -0
  350. package/dist/skills/opengame-trailer-plan/references/owner-eye.md +214 -0
  351. package/dist/skills/opengame-trailer-plan/references/reference-set.md +92 -0
  352. package/dist/skills/opengame-trailer-plan/references/review.md +141 -0
  353. package/dist/skills/opengame-trailer-plan/references/title.md +92 -0
  354. package/dist/skills/opengame-trailer-review-and-improve/CHANGELOG.md +5 -0
  355. package/dist/skills/opengame-trailer-review-and-improve/SKILL.md +145 -0
  356. package/dist/skills/opengame-trailer-review-and-improve/references/archetypes.md +113 -0
  357. package/dist/skills/opengame-trailer-review-and-improve/references/briefs.md +273 -0
  358. package/dist/skills/opengame-trailer-review-and-improve/references/core.md +129 -0
  359. package/dist/skills/opengame-trailer-review-and-improve/references/craft.md +299 -0
  360. package/dist/skills/opengame-trailer-review-and-improve/references/edge-cases.md +27 -0
  361. package/dist/skills/opengame-trailer-review-and-improve/references/filmable.md +244 -0
  362. package/dist/skills/opengame-trailer-review-and-improve/references/finishing.md +126 -0
  363. package/dist/skills/opengame-trailer-review-and-improve/references/gates.md +184 -0
  364. package/dist/skills/opengame-trailer-review-and-improve/references/measure.md +282 -0
  365. package/dist/skills/opengame-trailer-review-and-improve/references/music.md +148 -0
  366. package/dist/skills/opengame-trailer-review-and-improve/references/og-clash-record.md +133 -0
  367. package/dist/skills/opengame-trailer-review-and-improve/references/ops.md +154 -0
  368. package/dist/skills/opengame-trailer-review-and-improve/references/owner-eye.md +214 -0
  369. package/dist/skills/opengame-trailer-review-and-improve/references/review.md +141 -0
  370. package/dist/skills/opengame-trailer-review-and-improve/references/title.md +92 -0
  371. package/dist/skills/opengame-verification/CHANGELOG.md +11 -0
  372. package/dist/skills/opengame-verification/SKILL.md +113 -48
  373. package/dist/skills/opengame-weapons/CHANGELOG.md +11 -0
  374. package/dist/skills/opengame-weapons/SKILL.md +129 -55
  375. package/dist/skills/opengame-weapons/references/handling-numbers.md +42 -0
  376. package/dist/skills/opengame-weapons/references/reloads-and-hands.md +96 -0
  377. package/dist/skills/opengame-weapons/references/sights-and-scopes.md +98 -0
  378. package/dist/skills/opengame-weapons/references/threejs.md +71 -0
  379. package/dist/skills/opengame-world-physics/CHANGELOG.md +9 -0
  380. package/dist/skills/opengame-world-physics/SKILL.md +55 -24
  381. package/dist/skills/opengame-world-physics/references/measured-failures.md +28 -1
  382. package/package.json +1 -1
@@ -0,0 +1,170 @@
1
+ > Carried from `opengame-quality-bar` 1.1.0. Read it when: The bar a browser game is judged against on OpenGame, and how to judge your own game against it without fooling yourself. Covers nineteen cheap tells, the reference written as pass and fail statements, blunt scores, model judges that are anchored, paired and bound by the owner's decisions, class sweeps with an instrument per class, frozen approvals and a fresh reviewer. Use when the person asks is my game good enough, make it feel AAA, polish it, review my game, set up judges, or before a jam.
2
+
3
+
4
+ # The quality bar
5
+
6
+ Frontier models make scenes that look impressive in a still. They fail when the frame
7
+ stops and someone looks closely, or starts playing: a hero drawn as a capsule, a menu
8
+ that is a button on black, silence, a wall you walk through. A stranger notices in
9
+ seconds. This skill is the check for that, and the other `opengame-` craft skills are
10
+ how to fix what it finds.
11
+
12
+ This skill exists to close one gap. On OS the owner found about 93% of new defects and
13
+ the team's own checks about 7%, and on LC and OH the owner's hands and ears found what no
14
+ judge, bench or test did. Every rule below is a way of seeing the game the way the owner
15
+ sees it, before the owner has to.
16
+
17
+ ## How the rules are written
18
+
19
+ Every rule carries five tags, because a number that worked in one of our games is a
20
+ reference, not a law:
21
+
22
+ - **Invariant:** what must stay true in any game it applies to.
23
+ - **Technique:** one way to get there.
24
+ - **Reference configuration:** where specific values worked, and which game they came
25
+ from. DK is Der Koloss, MC Modern Claudefare, AW World at War and Astral War, SR
26
+ Skybreak and Smash Royale, ST the Studio that builds games on OpenGame, OC OG Clash,
27
+ a first-person wave shooter with a boss, built for WebGPU. OS is OG Snipers, a
28
+ first-person snipers-only multiplayer shooter on a wasm engine, with bots standing in
29
+ for players, drawn with WebGPU. LC is Last Contact, a first-person body-camera survival
30
+ horror wave shooter in three.js WebGPU, later with four-player lobbies. OH is OG Hero, a
31
+ painterly five-lane rhythm game in three.js WebGL2.
32
+ - **Applies when:** the genres, cameras and scales it fits.
33
+ - **Verify:** how to show it holds, by doing something rather than by reading code.
34
+
35
+ ## Rules
36
+
37
+ ### No cheap tells in what a stranger sees first
38
+ - **Invariant:** None of the cheap tells is visible on the route a new player takes: the title, the menu, the first minute of play, and the loading screen before them.
39
+ - **Technique:** Walk the list in [references/cheap-tells.md](references/cheap-tells.md) against captures from the real route, and fix them in its order: core loop first, then every generated asset in the game, then the front end, then a playtest.
40
+ - **Reference configuration:** ST holds a new game's first version until a review passes it, and that review names each tell it sees. DK, MC, AW and SR clear the first twelve; they are where the list came from. OC added the next four: one lazy-load change put low-poly guns in 6 places, a wall gun charged points while its model was missing, and real iPhones crash-looped during load after every emulated phone check had passed. OS and LC added the last three: OS's owner raised flicker, invisible walls and floating plants on five or more separate days each, and LC's owner raised arms and holds about 19 times.
41
+ - **Applies when:** Every game, every genre. A flat or minimal style changes what counts as a tell (flat light is then the style, not a defect), as long as the creator chose it.
42
+ - **Verify:** A reviewer with fresh context, given the captures and the list, marks each tell present or absent and cites the capture.
43
+
44
+ ### The bar is a frame from a named game
45
+ - **Invariant:** Before work starts on a screen, the target for it is a specific frame from a specific shipped game in the genre, or from the owner's own concept film, named in writing beside the concept art. The concept is the floor and the named frame is the bar.
46
+ - **Technique:** For each screen (combat, HUD, menus, boss, results) pick the genre leader and save one frame from it at the same camera and action as the concept. Put both on every comparison sheet. When the owner raises the bar mid-project, write the new reference down the same day and rescore against it.
47
+ - **Reference configuration:** OC used BO3 and BO6 for combat and the HUD, Fortnite for building and the third-person camera, Hades and Astro Bot for menus. The owner moved the bar from concept parity to "game of the year" on day 3; the critic then kept two scores, and from R67 to R77 the concept score went 4.4 to 5.3 while the reference-game score went 2.6 to 3.5. OC's HUD clusters took 25% of screen width; sized to the reference frame they take 13-15%. LC treated the owner's concept film as the floor ("the level of quality we need and ideally we even go beyond"); OH's judges scored the concept video as 10 on every criterion, and the owner stopped OH at a judge mean of 8.6.
48
+ - **Applies when:** Every game whose owner names a quality bar ("AAA", "like game X") or supplies a concept film. A game in its own invented style uses its best concept frame as both floor and bar.
49
+ - **Verify:** Every comparison sheet shows 3 panels per screen (concept, named reference frame, build) with the reference's name on it.
50
+
51
+ ### The reference becomes written pass and fail statements
52
+ - **Invariant:** Before a visual system is built, the owner's reference (a film, a concept video, a mood board, a named game) is written down as statements a stranger can mark pass or fail on a capture, measured where it can be, and every kind of visible thing in the game is graded against them.
53
+ - **Technique:** Cut the reference into contact sheets at a fixed interval plus a handful of hand-picked stills, and for each state you will build (rest, aim, a hit, night with the light off and on) find the reference frames of that same state before ruling on it. Write the look as statements about lens, exposure behaviour, light sources, palette, ground, hands and HUD, and measure what you can: luminance boxes at fixed places, the colour of the dark, the held item's share of the frame width. Name the layers the reference is built from (strokes, line, halftone, lens, motion) and build each as a technique, not a filter. Study a reference's method, never copy its code or art. Keep an inventory of every kind of visible thing graded placeholder, passable or final, and promote a row only with a capture beside the reference. Where the owner's newer words differ from the reference, the words win and the statement says so.
54
+ - **Reference configuration:** LC cut one frame every 5 s into 4 x 3 sheets plus 8 stills and wrote statements such as "rain only visible in the cone and the flash" and "with nothing lit the night is black, not dusk-blue"; a later pass cut 180 stills and measured the pistol at 4.7% of the frame width. That method made LC's darkness, pistol size and lens right, and missed the hands, reloads and fire, which fail in motion. A menu effect cost LC a day because the owner's "a bit trippy" was taken as a strength setting while nobody opened the reference. OH named its concept's layers and, once the owner sent a method video, rebuilt the world as "one long painting" (+0.54 judge mean in one pass), with every line written from scratch.
55
+ - **Applies when:** An owner supplies a film, trailer, concept video or mood board, or names a game to match.
56
+ - **Verify:** A reviewer who never read the code marks each statement pass or fail on a fresh capture, at the same state as the reference frame beside it.
57
+
58
+ ### Score by eye on side-by-side sheets, bluntly
59
+ - **Invariant:** Each screen gets a 0 to 10 score by eye against its concept and its reference frame, from fixed hero frames on a side-by-side sheet, and the bar the team scores against agrees with what the owner accepts and rejects. A pixel or colour metric can support a score and never closes an item on its own.
60
+ - **Technique:** Shoot the same hero frames every round. Have a critic who did not build the round mark every element in the concept present, wrong or missing, with its place in the frame. Cap the score at 5 until every element is present, and name the 3 biggest misses per frame. Ask the owner to score one sheet early and compare: a critic far above the owner is too soft and gets rewritten. Record the owner's acceptances as a state above the team's marks; when the owner passes what the bar fails, or fails what it passes, move the bar. Judge a moving thing as a whole act at speed before any still or crop.
61
+ - **Reference configuration:** OC's CIELAB metric improved from 16.56 to 14.68 while the frames were about 2/10, and it scored a black loading page 29.1; "parity 7.3" labels were later called meaningless. Once frames were scored by eye against one concept each, OC's critic rose from 3.5 to 6.4 over about 25 rounds; the owner rated comparable builds 1/10, 2/10 and 5/10, so OC treats a critic-owner gap over 2 points as a soft critic. OS's judges ran 1 to 2.5 points above the owner: frames they scored 2-3 got 1/10, and a map the owner called "probs a 5/10" got 7.4-7.8. LC's owner accepted all eight pistol reloads ("the pistol is pretty good reload already lets just do all the other guns") while LC's tester had marked one 6/7/5/7/4/5/5 against a pass of 9 on every point, and that ledger then stood at 0 of 73 for 27 hours.
62
+ - **Applies when:** Every game with concept art or a reference, from the first build the owner sees.
63
+ - **Verify:** Feed the scorer a death frame and a loading frame: both score near 0. Show the owner three items the bar passes and three it fails: the owner's calls match the bar's.
64
+
65
+ ### Model judges are noisy instruments: anchor them, pair them, and give them only what they can perceive
66
+ - **Invariant:** A judged score change counts only when the same judges saw old and new side by side, blind, in one sitting, against fixed anchors, with an unchanged control that held still. Judges score only what their evidence carries; sound, feel, input timing and motion at speed go to a person.
67
+ - **Technique:** One fixed rubric with the reference at the top and anchor frames from your own game at low, middle and high scores, the owner's own ratings binding. Three independent read-only judges, no two from one model family where you can, the median taken, a fourth added when they spread widely. Randomise old and new and keep the key outside the packet. One unchanged control per round, and a round whose control drifts is flagged. Count fixed defects by blind same-camera crop rechecks, never by a fresh sweep finding fewer. Tell judges to open the pictures, treat captions and commit messages as claims, return located defects with what would close each, and list what they could not see; the next packet carries that list. Act first on what two or more judges share, and check a lone judge's severe finding by eye. The setup is in [references/judges.md](references/judges.md).
68
+ - **Reference configuration:** OS judges drifted 0.5-3 points on identical frames (one map read 6.3, 8.1 and 5.9 with nothing changed) and about 1 item status in 5 flipped untouched; unpaired readings had the wrong sign on three maps where paired gave +1.4, +0.8 and +0.7; anchors cut spread from about 2.0 to 0.9 over 93 panels; blind rechecks reproduced 79 of 80 fixes while open re-sweeps swung from 26 to 1. Two OS judges from one family correlated at r = 0.81, so a median of three was two votes. OH's judges scored charts from text grids at 5.5-6.2 while the owner said 2-3/10, and only the owner found chart quality, the blank board and everything about sound. LC's single blind panel gave means of 4.65, 4.15 and 3.81 over 26 dimensions ("No frame I opened reaches the film") and sat about a point below LC's own playtest scores.
69
+ - **Applies when:** Any review or hill-climb that uses model judges.
70
+ - **Verify:** The control's drift and the paired spread print beside every score, and a panel on a frame the owner has rated lands within about 1 point of the owner.
71
+
72
+ ### Owner decisions travel in every judge brief, and outrank judges
73
+ - **Invariant:** No judge, critic, concept prompt or relayed note reverses a recorded owner decision or removes something the owner asked for, and the team makes no exception to an owner rule without the owner's yes. A conflicting critique becomes a question to the owner or a restyle, and one named reference sets each judged property.
74
+ - **Technique:** Keep a dated decisions file and paste it, word for word, into every judge, generator and worker brief under "must not reverse even if critics ask", with the spec facts judges must not penalise. Ask judges to score how well a choice is carried out, not the choice. Diff each panel's top fixes against the file before work starts. Relay the owner's words exactly, typos included. When the owner's verdict and the panel disagree, the owner's verdict stands.
75
+ - **Reference configuration:** Every OS judge seat ranked removing the default camo first or second, against the owner's own ask that new players spawn with that vivid camo; the lead removed it on judge advice and the owner restated it the same evening. Launchers added on one OS judge's ask cut a map's parity from 8.0 to 6.4. OS's judges never received the decisions list; the repair was the rule "Owner requests beat judges" and a prompt note to score how a choice is carried out. OH's judges erased the avatars' identity for about 18 h and recommended camera shake until the owner's rules sat in the rubric as "not a choice you may advise away"; one judge's taste flattened a pressed key the owner had asked for, and nobody noticed; OH kept slow songs as the studio's own exception to the owner's tempo rule, which cost a round trip.
76
+ - **Applies when:** Every critic, judge or image-generation loop on a game with an owner.
77
+ - **Verify:** Seed one frame that follows a decision and one that reverts it: the panel prefers the first.
78
+
79
+ ### One report becomes a numeric sweep, and every class the owner keeps finding gets an instrument
80
+ - **Invariant:** A defect the owner reports on one instance is rewritten as "no X may ever Y on any Z" with a measurable bar, and every member of the class meets it before the fix is reported. Any class the owner reports twice gets an automated check that sees it the way the owner does.
81
+ - **Technique:** Name the class (every gun, sight, camo, map, spawn, camera, weather, perk label, character), list every source that could produce the symptom, fix at the shared source where there is one, and sweep every member against a number on one sheet. Four or more instances of one category in one place mean one systemic cause. After each defect the owner finds that your checks missed, ask which instrument would have caught it; if none would, build one, prove it on a planted case, and let it fail on any rise above its last count. Report the class, its count and the members that failed, not the one example.
82
+ - **Reference configuration:** OS's owner found about 93% of new issues and the team's QA about 7%; flicker, floating plants, walk-through walls and stuck bots were each reported 3-7 times until each had a class audit, after which a glove-scale sweep held at "0.00 % (15,382 frames)". "Footsteps" took OS three rounds: rival steps, a turntable tick, then ambience beds generated with gravel in the prompt. OC fixed one optic on its own; a sweep of all 61 weapons, done only after the owner pushed, found 3 more defective guns, and a launch audit found 57 of 61 guns never captured by QA. OC's gun sheet is 61 guns by hip, aim and third person by base, upgrade and camo, 549 views, plus 152 acquisition cases. The owners: "how am i finding shit like this and u arent?" (OC); "whts the gap that all my feedback u cant find urself? fix that..." (OS).
83
+ - **Applies when:** Every defect report on a member of a set, and every class the owner raises more than once.
84
+ - **Verify:** The fix report links a sheet with every member marked pass or fail, the sweep reruns on the release build with the same zero, and each class the owner has found names its instrument.
85
+
86
+ ### Add what the concept has; remove what it lacks
87
+ - **Invariant:** Each round adds something the concept or reference frame has and the build lacks (a model, a light, a prop layer, an effect), or removes something neither has. Retuning one layer (sky, grade, fog) while the models, surfaces, lights and density still miss is not progress.
88
+ - **Technique:** Name one person who owns the frame as a picture and one who owns the renderer, on day one. Before tuning, list what the concept shows that the build does not, and check the renderer's limits (light count, model count) against what the scene needs. Try removing a cheap element before adding a new one.
89
+ - **Reference configuration:** OC spent hours grading before the owner called the build "made with a crayon"; the frame had no owner. The render audit that finally ran found a 32-light cap against 218 fire sites, and the content side of the fix is in `opengame-assets`. A flat cyan visor strip took 3 h 40 min and 3 private builds to remove, after which the owner called that visor perfect.
90
+ - **Applies when:** Every game being pushed toward a concept or reference, most of all in 3D.
91
+ - **Verify:** Each round's notes name what was added or removed and point at it on the sheet.
92
+
93
+ ### Change the method, not the polish, when progress stalls
94
+ - **Invariant:** When several passes in a row each move the score by little, or the owner rejects the same axis two or three times, the next pass changes the method (how the thing is made, what feeds it, who owns it) instead of polishing the current one again.
95
+ - **Technique:** Log each pass's score delta and the owner's verdict per axis. At a stall, find out how the reference itself was made (ask the owner, read a method walkthrough), rebuild one showcase on that method and score it beside the old one. Cut briefs to the top three items when the same reasons repeat, because broad passes touch many things lightly. Never let a loop work one sub-part for hours while the largest gap sits elsewhere.
96
+ - **Reference configuration:** OH's repaints gained +0.13, +0.20 and -0.07 in a row; rebuilt on the owner's method video, the world gained +0.54 in one pass. OH spent about 26.7 hours charting songs nobody on the team had written (nine owner complaints, ending at "2 or 3/10") before composing its own, and the first composed song drew "great chart" about 16 hours after the pivot. An 11-item OH pass fell where a 3-item pass rose. OS's overnight loop made about 60 builds on the fists while maps sat at 3-4/10: "Wait the whole night u worked on the fists only and thats where u got? It looks worse than when u stRted". OH treats about three passes under +0.3 as a stall.
97
+ - **Applies when:** Any long hill-climb against a concept, a reference or owner notes.
98
+ - **Verify:** Before each new round, a decision note shows the last three deltas and the owner's latest verdict, and names the method change when they show a stall.
99
+
100
+ ### Once free concepts stall, targets are paint-overs of the build's own frames
101
+ - **Invariant:** A screen's visual target shares the build's camera, layout, HUD and state, so every difference between target and build can be acted on, and progress is scored on captures of the build, never on the target.
102
+ - **Technique:** Register a camera (eye, forward, field of view), capture the build there, and have an image model change only the rendering ("KEEP EXACTLY camera/layout/HUD; CHANGE ONLY THE RENDERING"), on the person's own image tool and with their yes on any cost. Pin each target by hash, reject paint-overs that invent geometry or content the camera cannot see, and have each list what it gained and what the build still misses. Move what the image model cannot draw into an engine spec. Plot the build's score and the target's score on separate lines: a paint-over score measures the generator.
103
+ - **Reference configuration:** OS's 293 free concepts stalled at a best of 8.0 against an 8.5 bar ("This is a ceiling, not slow progress"), while paint-overs of captured frames scored higher (7.3 against 6.95) and gave per-frame "still missing" lists. Failures: an invented titan, and a river "because the prompt asked for a river that this camera can't see"; writing "NO water anywhere in this frame" lifted same-scene agreement from 2/4 to 4/4. OS's paint-over medians sat at 7.5 every round while its build panel sat at 3-4. OS made 483 paint-overs at no cost on a subscription image tool.
104
+ - **Applies when:** Hill-climbing environment or screen visuals against generated targets, once free concepts stop improving.
105
+ - **Verify:** Build and paint-over scores are plotted per round on separate lines, and the next capture from each pinned camera shows the items its paint-over listed.
106
+
107
+ ### A stranger wants a second minute
108
+ - **Invariant:** Someone who has never seen the game, starting from the real menu with no explanation, understands what to do, does it, and wants to keep going.
109
+ - **Technique:** The ninety-second stranger test: open the published or preview build cold, go through the real menu (never a developer shortcut past it), play for ninety seconds, and write down every moment of confusion, waiting or boredom.
110
+ - **Reference configuration:** The bar our originals are held to: a person clicks one game, plays it for ninety seconds and forms an opinion about the studio. A game that boots to black has already answered them.
111
+ - **Applies when:** Every game.
112
+ - **Verify:** Record the ninety seconds as video or a frame sequence with input. A pass names the first action taken, when it happened, and the first consequence the player saw.
113
+
114
+ ### The builder never grades its own work
115
+ - **Invariant:** The verdict on a piece of work comes from someone who did not build it and who sees the artifact, not the builder's explanation of it.
116
+ - **Technique:** Give a separate agent with fresh context the bar, the captures and the running build, never your summary. It marks every area met, gap or unverified, and names the single largest gap. Fix that gap, then ask again. The prompt is in [references/review-prompt.md](references/review-prompt.md).
117
+ - **Reference configuration:** ST runs an independent reviewer on every first version. In our own repository, a second review of one day's fixes found the first review's fixes carried the same defect classes they were fixing. OC's first independent review after integration found 30 ranked defects that every self-check had passed, and later an adversarial review of each delivery found a real bug in every one. About 30 LC lead claims were overturned when the director opened or played them.
118
+ - **Applies when:** Every game, at every checkpoint the person will see.
119
+ - **Verify:** The review names a single largest gap, cites a capture for each verdict, and marks anything it could not observe as unverified rather than met.
120
+
121
+ ### Everything made for the game is in the game
122
+ - **Invariant:** Every model, rig, sound, clip and image generated for this game is loaded and seen or heard on the real route, and nothing a stand-in occupies has a finished asset waiting unused. An asset counts once someone has seen or heard it in the running game.
123
+ - **Technique:** Keep a list of every generated asset with the call site that uses it. Grep for each file name in the game's source; play to where it appears.
124
+ - **Reference configuration:** ST fails a first version on any generated work left unused, and lists those first in every playtest report. Our originals once shipped thirty models into eight games where only two games loaded a model at all. OC paid for mocap attack clips that sat in the zombie files unused, generated props were flagged because "nobody has looked at them in-engine", and the title scene's cigar clip did not play until R102, after the owner asked "where is the cigar smoking?" 3 times. LC had reload sounds for every gun that nobody on the team had heard.
125
+ - **Applies when:** Every game that generated assets.
126
+ - **Verify:** For each asset, a source line that loads it and a capture or recording where it appears in play.
127
+
128
+ ### Judge the frame people play, and make a default only when it is right
129
+ - **Invariant:** Quality is judged from the player's camera, at play distance and the player's settings and window, while moving, not from an editor view or a hero still. A change becomes the default only when it is right from that view, never because it beats the placeholder it replaces, and never while its maker's own report names a failure from that view.
130
+ - **Technique:** Capture from the real camera during real input: a still one second after Play, a pan, a sprint, a hit. Compare against the concept at the same camera and action, on full-size frames, never reduced ones. Block a default while the lead's own notes name a failure from the player's view.
131
+ - **Reference configuration:** ST compares player-camera captures against each selected concept view; a canvas-only capture cannot prove a HUD exists, so its captures include the DOM. OC's owner moved from judging stills to playing on day 2, and the first play-test list that evening had 50 items. LC's baked fire was made default as "still better than the old pale ball", judged on 896 x 430 captures and from the side, while its lead had written that from the hands it was "ONE UPRIGHT COLUMN"; two hours later the owner, from the hands at Medium: "literally looks so bad".
132
+ - **Applies when:** Every game with a camera, and every default change to something the player sees.
133
+ - **Verify:** The captures are timestamped, include the HUD and are taken through the real menu path, and each default's commit cites a full-size frame from the player's view at the player's settings.
134
+
135
+ ### Approved quality is frozen: never traded, never made worse
136
+ - **Invariant:** Art, effects or fidelity the owner has approved or accepted are not cut, lowered or swapped for a budget, a device, a deadline or a judge's advice unless the owner agrees to that specific trade. The approved frame and commit are the target no later pass may make worse, and a regression is fixed by exact restore first.
137
+ - **Technique:** Keep an approved-looks record (frame, commit, the owner's words, date) and an accepted list. A change to an approved surface ships with a board of approved, current and proposed at one zoom and must beat both the approval and the reference. When a budget is missed, optimise residency, streaming and shader cost first; if a cut is still needed, show the before, the after and the measured saving, and wait for a yes. On "worse than round N", A/B a straight revert (asset, transforms and shader together) against the fix-forward before anything else. Never let a device tier or fallback lower quality without saying so on screen.
138
+ - **Reference configuration:** OC agents cut camos to save memory and the owner reversed it ("do NOT cut down the sick camos"); phone builds shipped fewer enemies, smaller textures and half the sky without asking, all reverted at R179, and the owner overrode a blanket "bytes only go down" rule once the tradeoff was shown. OS's approved scope was broken by a parity pass and then by a rebuild "from measurements" ("it used to be perfect this scope this is what we need to bring it back to"), and only restoring the approved commit fixed it; an arm regression was fixed forward until "the arms and hands looked MUCH better in l169". OS's "im fine with the fists/hands view as is no need to spned more time on that piece" ended hands work at a judge score of 3.7: an acceptance is a stop.
139
+ - **Applies when:** Any surface the owner has praised, accepted or used as a comparison, and every budget or device fallback.
140
+ - **Verify:** Each quality cut in the change log links the owner's approval, every change to an approved surface has its board, and no later owner note says "it used to be".
141
+
142
+ ### Every picture sent to the owner is held to the ship bar
143
+ - **Invariant:** No image in a progress report shows a flaw its caption does not name, because the owner reviews every picture sent, whatever it was meant to show.
144
+ - **Technique:** Before posting a frame, list its visible defects at full size. Prefer frames from the current default build at the player's view over crops of working composites. Put the images themselves in the message, never file paths. A dark frame brightened to be read says so on the image.
145
+ - **Reference configuration:** LC's first two owner screenshots were crops of the director's own composites, posted to show materials and a playtest bug; the owner rejected the hold and the resolution in them instead. OS's owner: "when you send comparison sheets i need u to send the actual iamges here in the chat! i cant see anything". LC stamped every brightened tile with its gain, so no lifted picture was mistaken for the player's.
146
+ - **Applies when:** Every status update or report with images.
147
+ - **Verify:** Each posted image carries a defect list or "none seen at full size", and none is a path.
148
+
149
+ ### Report what was observed
150
+ - **Invariant:** Every claim about the game is marked observed pass, observed failure, unverified or not applicable, and a source read is never reported as an observation.
151
+ - **Technique:** A checklist of at most six rows per checkpoint: check, route, action and expected result, observed result, evidence, next fix. Carry unresolved rows forward. Grade each claim by its evidence: seen in the build the owner plays, seen in someone else's capture and not opened by you, by count, by test only, or behind a flag and unseen. Read every reported commit hash back from the repository.
152
+ - **Reference configuration:** ST's checkpoint table keeps "unverified" as a first-class status so an untested claim cannot read as a pass. OC's ask register marked each ask fixed, partial, not fixed or could not check, with frame proof, after an audit found 0 of 67 inherited asks closed with on-screen proof. LC graded every claim this way after reporting work as "in" on the strength of commit messages ("I should have said 'written, unseen'"). OH found one fabricated commit hash in its own reports by reading the repository.
153
+ - **Applies when:** Every report to the person, every submission note.
154
+ - **Verify:** Every row marked pass links a capture, clip time or test output.
155
+
156
+ ## The other craft skills
157
+
158
+ Look: `opengame-look`. Light: `opengame-lighting`. Collision and movement:
159
+ `opengame-world-physics`. Geometry defects: `opengame-scene-audit`. Levels, spawns and
160
+ lanes: `opengame-maps`. Feel: `opengame-game-feel`. Characters: `opengame-characters`.
161
+ Sound: `opengame-audio`. Menus and HUD: `opengame-front-end`. Frame time:
162
+ `opengame-performance`. Generated assets: `opengame-assets`. Systems: `opengame-depth`.
163
+ Proving it works: `opengame-verification`. Everyone can play: `opengame-accessibility`.
164
+ First-person weapons and viewmodels: `opengame-weapons`. Enemy AI, navigation, hordes,
165
+ bosses and deaths: `opengame-enemies`. Boot, streaming, caching and honest progress:
166
+ `opengame-loading`. Touch, phone memory and real Safari: `opengame-mobile`. Crash-free
167
+ play, budgets for every pool and error reporting: `opengame-stability`. Lobbies and
168
+ netcode: `opengame-multiplayer`. Rhythm games: `opengame-rhythm`. Filming and
169
+ screenshots with a fixed clock, a clean feed and a free camera: `opengame-capture`.
170
+ Running the build loop with agents over many rounds: `opengame-production`.
@@ -0,0 +1,43 @@
1
+ # The nineteen cheap tells
2
+
3
+ Each is something a stranger notices in seconds, drawn from what our own games had to
4
+ fix (DK, MC, AW, SR, OC, OS, LC) and what the Studio's reviewer finds in first versions
5
+ (ST). For each: what it looks like, how to see it, and which skill fixes it. Tells 13 to
6
+ 16 came from OC, a large 3D game played on desktops and phones. Tells 17 to 19 came from
7
+ OS and LC: OS's owner raised flicker and collision on five or more separate days each, and
8
+ LC's owner raised arms and holds about 19 times.
9
+
10
+ | # | Tell | What it looks like | How to see it | Fix with |
11
+ | --- | --- | --- | --- | --- |
12
+ | 1 | Primitive stand-ins | A hero, creature, vehicle, item or landmark drawn as a box, sphere, capsule or cone where a designed thing belongs | Grep for `BoxGeometry`, `SphereGeometry`, `CapsuleGeometry` under item and actor names; look at play distance | `opengame-assets`, `opengame-characters` |
13
+ | 2 | Default materials | Flat colour, roughness 1 or one roughness everywhere, no maps, metal that reflects nothing | Grep `MeshStandardMaterial({ color` with no maps; look at a surface while moving | `opengame-look` |
14
+ | 3 | Flat light | Shadowed surfaces nearly as bright as lit ones, no cast or contact shadows, no warm and cool contrast | Compare a lit wall and its shadowed neighbour; isolate the ambient term | `opengame-lighting` |
15
+ | 4 | Missing post | No tone curve or grade, bloom on lit walls, a milky horizon, a frame flatter than the concept | Capture with the grade off and on; look at the horizon | `opengame-look` |
16
+ | 5 | An empty world | A ground plane with props on a grid, nothing overhead, no wear, props floating or intersecting | Run the scene audit; pan slowly at a grazing angle | `opengame-depth`, `opengame-scene-audit` |
17
+ | 6 | Placeholder UI | Browser-default fonts, debug text, unstyled boxes, emoji or Unicode glyphs as icons, a title that is a button on black | Load the menu cold; inspect computed fonts | `opengame-front-end` |
18
+ | 7 | Silence, thin or harsh sound | No music or ambience, one sample for every action, clipping when sounds overlap, a sound that is harsh, late or louder than the shot it should sit under | Play with headphones through a busy minute; a person listens, since no agent can | `opengame-audio` |
19
+ | 8 | No hit feedback | A hit with no marker, flash, sound or reaction; a kill with no beat of its own | Hit something and watch the next half second frame by frame | `opengame-game-feel` |
20
+ | 9 | A dead camera | Rigid follow, no easing, no response to sprint or impact, shake that never decays, a camera inside walls | Sprint, land, take a hit, back into a wall | `opengame-game-feel` |
21
+ | 10 | Stiff animation | Sliding feet, clips that pop with no crossfade, no hit reaction or death, a T-pose anywhere | Watch feet at slow, normal and sprint speed; kill something | `opengame-characters` |
22
+ | 11 | Hitches | A stall on the first shot or effect (shader compile), frame-time spikes from allocation | Record frame times through first combat; read p95 and p99 | `opengame-performance` |
23
+ | 12 | Generated work left out | A model, rig or clip made for this game that the game never loads, while a stand-in sits in its place | List generated files; grep each name in source | `opengame-quality-bar` |
24
+ | 13 | Stand-ins while loading | A low-poly copy, a silhouette, a wrong variant or a half-built mesh shown in a menu, a pickup, a reward spin or a customisation preview while the real asset loads | Throttle the network, open every menu and acquisition path, and capture the first frame each item appears | `opengame-loading`, `opengame-assets` |
25
+ | 14 | A black or white held item | The gun or hands in first person read as a black shape or a blown-out white one, in dark corners or after a shot | Capture the held item in the darkest and brightest spots of the map and on the frame after firing | `opengame-weapons`, `opengame-lighting` |
26
+ | 15 | Stuck enemies | Enemies or allies pressed against a wall, pacing on a ledge, frozen in the air, or a round that never ends because one is stranded | Stand at every spot the player can reach and count how many enemies arrive | `opengame-enemies` |
27
+ | 16 | Crash or reload during load | The tab reloads, the browser says a problem repeatedly occurred, or the game drops back to the menu near the end of the bar | Cold loads on a real phone in Safari and on a throttled desktop, several in a row | `opengame-stability`, `opengame-loading`, `opengame-mobile` |
28
+ | 17 | Flicker in motion | A surface that shimmers, crawls, stripes or swaps while the camera turns, strafes or zooms, though every still looks clean | Film a fixed still, turn, strafe and scoped-pan path on the game clock and step through it frame by frame | `opengame-scene-audit`, `opengame-look` |
29
+ | 18 | The picture and the collision disagree | An invisible wall or bullet stop, a solid you walk through, a doorway you cannot pass, a spawn inside a rock or facing a wall | Walk the real controller through every opening and from every spawn; shoot every surface that looks solid | `opengame-world-physics`, `opengame-maps` |
30
+ | 19 | A body no person has | First-person arms floating or cut off, a hold no hand makes, wrists bent past a natural range, a body that squashes or slides, seen at speed in the player's own picture | Film the held item at rest, aiming, reloading and turning, and watch it at speed and at quarter speed from the player's camera | `opengame-characters`, `opengame-weapons` |
31
+
32
+ ## The order to fix them in
33
+
34
+ When time is short, in this order, because each step makes the next one worth doing:
35
+
36
+ 1. The game loads and plays without crashing on the target devices, and the core loop
37
+ and its flagship moment work and feel right.
38
+ 2. Every generated asset is in the game where it belongs, and no stand-in shows while
39
+ it loads.
40
+ 3. The world holds together in motion: no flicker, no invisible walls, no starts inside
41
+ geometry.
42
+ 4. The title, menus, pause and results are real screens in the game's own language.
43
+ 5. A playtest from the real menu, reviewed by somebody who did not build it.
@@ -0,0 +1,95 @@
1
+ # Running a council of model judges
2
+
3
+ How to set up model judges so that a score change is bigger than their noise, and what
4
+ they cannot do. From OS's council (879 judge tasks over about 71 hours), OH's three-judge
5
+ panel (25 rounds on one rubric) and LC's single blind panel. Judges are instruments: they
6
+ need calibrating like any other, and their output is a defect list first and a number
7
+ second.
8
+
9
+ ## What goes in the packet
10
+
11
+ - **Fixed frames per screen**, from the same cameras every round, captured through the
12
+ real menu from a frozen build (`opengame-capture` says how). Stack them into contact
13
+ sheets: OS judges given about 100 single images stalled on image-read limits (one ran
14
+ 4.5 hours and 2,659 reads with no scorecard), and 12 stacked sheets cut a run to 3-4
15
+ minutes.
16
+ - **The reference frames of the same state** beside the build's frames. LC's prompt:
17
+ compare "the game's frames with the concept film's frames of the same kind of moment".
18
+ - **Short clips** for anything that moves. A still cannot score flicker, shake, pacing or
19
+ a transition, and still panels sat at 3-4 on OS for about 135 builds while the owner's
20
+ notes were about motion.
21
+ - **The owner's decisions file, word for word**, under "must not reverse even if critics ask",
22
+ plus the spec facts judges must not penalise.
23
+ - **What the last panel said it could not see**, so the next packet answers it.
24
+ - **No claims of progress**: no round labels, no captions that say what improved. OS's
25
+ stricter briefs said "Do not infer improvement from round labels" and "unknown evidence
26
+ => unverified, never a score".
27
+
28
+ ## The rubric
29
+
30
+ - One rubric for the whole project, so rounds compare. OH used ten criteria from 0 to 10
31
+ with the concept video as 10, the last being whether it "reads as a AAA, big-budget game
32
+ from a real studio"; LC used 26 dimensions, from fear and photorealism to load and
33
+ stability.
34
+ - Score meanings LC wrote down: 10 is indistinguishable from a first-party console release
35
+ in that dimension, and the judge looked for a defect and could not name one a paying
36
+ player would notice; 8-9 shippable with named shortfalls; 6-7 competent, clearly not AAA;
37
+ 3-5 prototype; 0-2 missing. A 10 needs a sentence saying what was checked. LC scored no
38
+ evidence as 0, but any fixed score for an unseen item makes the mean track coverage: an
39
+ OS panel read 1.12/10 because 16 of 17 maps were never framed. Mark a dimension with no
40
+ evidence unverified, leave it out of the mean, and report coverage beside the score.
41
+ - Anchor frames from your own game at 3, 5 and 7, and the reference at 9 when it is a floor
42
+ the owner wants exceeded or at 10 when it is the bar, with any rating the owner has given
43
+ an anchor binding (OS: visual spread fell from 1.8 to 0.6).
44
+ - A vision probe before scoring, so a judge that cannot read images is caught before it scores. OS asked
45
+ for "line 1 'VISION: yes'" and then "the three most prominent colours".
46
+
47
+ ## Running a round
48
+
49
+ - Three judges, no two from one model family where you can (two OS judges from one family
50
+ correlated at r = 0.81, so a median of three was two votes), each a fresh read-only task
51
+ that sees nothing of the others. Never fork a judge from the build conversation: some OS
52
+ "blind" judges were forked with the whole owner conversation. Take the median, and add a fourth judge when the spread is over 2 (OS).
53
+ For the defect list, the lowest score per dimension counts (LC).
54
+ - Paired, blind, same-sitting A/B: old and new in one packet, order randomised, the key
55
+ kept outside the packet. Unpaired OS readings had the wrong sign on three maps where
56
+ paired readings showed gains.
57
+ - One unchanged control per round. On OS a control drift over 0.5 flagged the round, and 2
58
+ of 3 trains were flagged.
59
+ - Moves under 0.5 count as nothing: OS's three judges agreed on the direction of change in
60
+ 35 of 115 map-rounds.
61
+ - Output as JSON per dimension: score, what was checked, defects with where, what and what
62
+ would close the gap, and which area owns it, plus a short verdict (LC capped it at 700
63
+ words). Each judge checks that every path it cites exists.
64
+ - Defects as located boxes, kept when two or more judges mark the same place: 704 OS
65
+ defects at about 80% precision.
66
+ - Fixed defects counted by blind same-camera crop rechecks (79 of 80 reproduced on OS),
67
+ never by a new sweep that happens to find fewer.
68
+
69
+ ## What judges cannot do
70
+
71
+ - **Hear.** OH's judges saw silent clips only; every sound defect in OH and LC was found
72
+ by the owner.
73
+ - **Feel.** No judge pressed a key. OH's judges scored charts from text grids while the
74
+ owner played them and said "2 or 3/10".
75
+ - **Watch at speed.** A judge reads frames in order; a pop or a whirl that lives between
76
+ frames goes unseen.
77
+ - **Keep the owner's taste unprompted.** OH's judges erased the avatars' faces and asked
78
+ for camera shake; OS's judges removed a default the owner had chosen. Personas do not
79
+ help: OS's three seat means were 7.04, 6.98 and 7.03.
80
+ - **Resist easy wins.** Defect counts reward deletion (a clipping plant was simply
81
+ removed), and judges overcorrect each other: an OS HUD was "30-40% too big", then
82
+ "shrunk too far", and a scope lens went 648, 961, then 886 px.
83
+
84
+ So report judge scores as judge scores, keep an owner-play log beside the judge log, and
85
+ never present a proxy score to the owner as progress.
86
+
87
+ ## Cost and cadence
88
+
89
+ - OH: about 25 minutes of capture, 2-11 of judging and 10-30 of briefing, so a round every
90
+ 1.5 to 2.5 working hours. Judges ran on subscription pools; never buy judging on paid
91
+ API credit without the person's yes on the cost.
92
+ - Run a panel early, for the defect list, and again when you believe every dimension is
93
+ close. Once the last panel's top fixes have landed, judge only what changed (OS).
94
+ - Never gate publishing on a bar no build can reach: an OS panel at 1.12/10 held the
95
+ playable preview on one build for about 8 hours 40 minutes while 28 newer rounds waited.
@@ -0,0 +1,32 @@
1
+ # A prompt for the fresh-context reviewer
2
+
3
+ Give this to a separate agent that has not seen the build conversation. Attach the
4
+ captures, the build's address and the owner's decisions file, never a summary of what
5
+ you changed.
6
+
7
+ > You are reviewing a browser game against a quality bar. You did not build it and you
8
+ > owe its builder nothing. Use only what you can observe in the attached captures and by
9
+ > playing the build at the address given, starting from its real menu. Open the pictures
10
+ > at full size. Captions, commit messages and status notes are claims, not evidence.
11
+ >
12
+ > 1. For each of the nineteen cheap tells (listed below), say present or absent and cite
13
+ > the capture or the moment in play. If you could not observe it, say unverified.
14
+ > 2. Play for ninety seconds as a stranger. Write the first action you took, when you
15
+ > took it, and the first consequence you saw. Write every moment you were confused,
16
+ > waiting or bored.
17
+ > 3. For each area (control, look, light, sound, feedback, camera, animation, menus,
18
+ > performance, world), mark met, gap, unverified or not applicable, with evidence
19
+ > or a reason it does not apply to this game.
20
+ > 4. Name the single largest gap: the one change that would most raise how this game
21
+ > reads to a stranger. Be specific about where and what.
22
+ > 5. List what you could not perceive from what you were given (sound, input feel,
23
+ > motion at speed, a screen nobody captured), so the next review can be given it.
24
+ >
25
+ > Do not grade effort or intent. Do not accept an explanation in place of an
26
+ > observation. A "fine" that you cannot point at is unverified. The owner's decisions
27
+ > below are not yours to reverse: judge how well each is carried out, not the choice.
28
+
29
+ Then paste the decisions file and the table from [cheap-tells.md](cheap-tells.md).
30
+
31
+ Fix the single largest gap, capture again, and ask a fresh reviewer, not the same one:
32
+ a reviewer who has seen the previous round starts to see what they expect.
@@ -0,0 +1,134 @@
1
+ > Carried from `opengame-rhythm` 1.0.0. Read it when: Make a rhythm or music game whose charts feel right to play. Covers owning the music so charts match it, nested levels, hand limits, keys that follow pitch, chords and holds, a setlist career, the audio hardware clock, hit windows, calibration, a board nothing covers, star power, mix timing and checking every chart through the game. Use when the person says the notes feel random, Expert is impossible, hits feel late, or asks for a rhythm, music, lane or beat game.
2
+
3
+
4
+ # Rhythm games
5
+
6
+ A rhythm game lives or dies on one feeling: the keys you press are the music you hear.
7
+ Every other part, the painted world, the crowd, the effects, is judged after that, and
8
+ a player who feels "random keys at random times" stops in the first song. OH spent most
9
+ of a day and a half learning that the hard part is not timing, which was solved early,
10
+ but choosing which notes to chart, and that the way to choose them is to own the notes.
11
+
12
+ Rules carry five tags; `opengame-quality-bar` explains them and the game codes. The
13
+ measured genre densities, the reduction weights, the hand gates and the key mapping are
14
+ in [references/charting.md](references/charting.md). The song clock, calibration, the
15
+ count-in and codec delay are in [references/timing.md](references/timing.md). Mixing and
16
+ mastering in general are in `opengame-audio`; hit feedback in general is in
17
+ `opengame-game-feel`.
18
+
19
+ **Never fetch or copy a commercial game's audio or charts.** Calibrate against aggregate
20
+ statistics (notes a second, gaps, sustain share) and against music you have the rights to.
21
+
22
+ ## Rules
23
+
24
+ ### Chart only music whose notes you own
25
+ - **Invariant:** Every note on the board is a note of a written featured part, played at its measured sounding time. Nothing is transcribed from a mixed recording, and the game plays the chart exactly as written.
26
+ - **Technique:** Compose songs as data (a small score language with bar checks and phrase marks), render the featured stem on its own, chart from its notes, and keep the game's chart loader free of any listening, moving, filtering or thinning. When the music must come from elsewhere, get the stems or the score from whoever owns it rather than separating a mix.
27
+ - **Reference configuration:** OH charted existing songs for 26.7 h across six extraction rounds: whole-mix onsets ("completely random keys at random times"), a melody tracker on separated stems (precision 0.50, recall 0.61), regression and boosted trees on 18 note features (best 0.63 / 0.57), and a large transcription model (0.22 / 0.26). Timing was never the fault (4 ms median); choosing which notes was, and in strummed songs "86% of what our system calls 'the melody' there is just the top note of each chord". The owner rated the best "2 or 3/10". The owner then asked for composed music, and the first composed song, charted as written, drew "great chart" 16 h later. Featured-stem onset p95 was 4.4 ms against a 10 ms gate.
28
+ - **Applies when:** Any rhythm game whose content you can author. A game built on licensed tracks needs the tracks' stems or scores from the rights holder.
29
+ - **Verify:** A checker matches every chart note to a score note at every level, and the owner plays one song on the hardest level before the rest are made.
30
+
31
+ ### The hardest level is the part as heard, and easier levels are nested subsets
32
+ - **Invariant:** Every note on a lower level sits on a note of the level above, at its time and key; a chord may lose a key on the way down and never gains one; a repeated phrase is reduced the same way each time; density rises with every level on every song.
33
+ - **Technique:** Score each note by metric position, length, syncopation, phrase edges and melodic turning points, keep the top share per phrase, and cap notes a second per phrase. Fit the keep shares by feeding published hardest-level charts to your thinning and comparing what you keep with what the published easier levels keep, counts only.
34
+ - **Reference configuration:** OH keeps Hard 0.86 (cap 7.5 notes a second in a phrase), Medium 0.66 (cap 5.0, chords at most 12%), Easy 0.52 (cap 2.8, no chords, at most 86 to 88% of Medium). Published levels in the genre keep about 85, 61 and 41% of Expert on dense songs; OH's first rules kept 60, 47 and 32% until refitted to 82, 66 and 44%. Lower levels in the genre sit on Expert note times 99.5% of the time.
35
+ - **Applies when:** Any chart with more than one difficulty.
36
+ - **Verify:** A subset assertion over the whole catalogue, and a per-song table where notes a second rise Easy, Medium, Hard, Expert.
37
+
38
+ ### Hands set the limits, and every level uses every lane
39
+ - **Invariant:** No level asks for more than a hand can play, every level uses all the lanes, and difficulty comes from density, spacing and chords rather than from removing lanes.
40
+ - **Technique:** Gate same-key repeats, any two notes and the busiest second per level, and thin an unbroken run of fast notes into a roll and then into beats. Read the owner's "impossible" as a gate that is missing, not as a request for fewer notes everywhere.
41
+ - **Reference configuration:** OH: the same key at least 150, 150, 190 and 300 ms apart (Expert, Hard, Medium, Easy); any two notes at least 85 ms apart; the busiest second at most 10 notes; Easy at most 2.6 notes a second with no chords. Genre medians: 1.61, 2.44, 3.40 and 3.98 notes a second, median gaps 480, 323, 214 and 189 ms, busiest seconds 4, 5, 8 and 10. Stem charts at 12 notes a second with chords everywhere drew "this is fucking impossilbe"; melody-only charts below the genre's Easy density then drew "missing so many notes".
42
+ - **Applies when:** Lane games played on keys or pads without a strum bar.
43
+ - **Verify:** The checker names the bar of every collision, and the owner plays the hardest song on the hardest level without calling it impossible.
44
+
45
+ ### Keys follow pitch, and phrases come back on the same keys
46
+ - **Invariant:** Within a phrase a higher note is further right and the same note is the same key; a returning phrase returns on the same keys; fast runs are rolls that never repeat a key.
47
+ - **Technique:** Give up to as many pitches as there are lanes their own key per phrase, spaced by interval; split wider phrases where the tune steps least; move the hand only at marked phrase points. Chart the lead line the player can hear, never the drums.
48
+ - **Reference configuration:** OH: a step or a third lands next door, a fourth to a sixth skips one lane, and a `/` in the score marks where the hand may move. Genre lane moves on Expert run about 28, 50, 14 and 7% for 0, 1, 2 and 3 or more lanes. The owner: "we dont need to have the kick durm or snare ... its for the main tune/instrument/lead". An arpeggio "up and down" was charted off until it became a roll that follows its contour.
49
+ - **Applies when:** Assigning lanes from pitched notes.
50
+ - **Verify:** Pitch to key is a function within each phrase, checked by script, and every arpeggio is read on a grid at every level.
51
+
52
+ ### Chords and holds come from the music
53
+ - **Invariant:** A chord appears only where two notes are heard, a hold only on a note that rings, chord width shows the interval, and an easier level never adds either.
54
+ - **Technique:** Build accent chords from a tone the band already holds, a third to a sixth below the melody, on whole beats, and group them per repeated phrase so repeats chord alike. End every hold before the next note.
55
+ - **Reference configuration:** OH chord targets run from 8% of featured notes at tier 1 to 28% at tier 8; holds at 1.5 beats and 0.5 s, or 1 beat and 0.8 s, ending 40 ms before the next note. Grouping lifted alike-chorded repeats from 45 to 88%. The owner, after the first good chart: "dont be so stingy on the chords too". The shipped catalogue still had median Expert chord shares of 0.7 to 4.4% per tier, an ask left open because re-makes were queued and never ran.
56
+ - **Applies when:** Writing or thinning charts.
57
+ - **Verify:** A chord-share table per tier against its target, and a hold that stays visible and catchable for its whole length in a capture.
58
+
59
+ ### The setlist is a career and every song earns its place
60
+ - **Invariant:** One list of rising difficulty in tiers, each tier ending on a closer, placed by measured density; no song is boring, too slow or too long; a preview starts at the song's best part.
61
+ - **Technique:** Write the setlist before the songs. Move a song down a tier rather than pad its chart. Declare each song's highlight and preview from it. Give the music a hook that comes back, and keep the lead in a comfortable register.
62
+ - **Reference configuration:** OH: 8 tiers of 5 hand-written songs, then generators, 169 songs over 17 styles; Expert at 2.2 to 2.9 notes a second at tier 1 rising to 5.1 to 6.3 at tier 8; at least 118 bpm with no exceptions (three slower songs kept as the studio's own exception were removed by the owner); 2:00 to 2:30 a song; previews from the first chorus after a 320 ms settle with a 0.18 s fade. Leads pitched an octave high lost 0.4 to 0.9 of a listener model's score. The owner: "we dont need 4000 songs ... we just need a good maount of songs with AMAZING charts".
63
+ - **Applies when:** Any game with a song catalogue.
64
+ - **Verify:** The index orders with no gaps, tiers ascend by measured density, and a person listens to five previews.
65
+
66
+ ### The song clock is the audio hardware's clock
67
+ - **Invariant:** Judging and drawing read song time from the audio output, never from frame time, and each key is judged at its own event timestamp, so a slow frame cannot make a hit late.
68
+ - **Technique:** Start one decoded buffer at a scheduled context time, map `getOutputTimestamp()` to song time with an output-latency fallback, predict it forward each frame and glide, and convert `event.timeStamp` inside the key handler. A note is marked missed by the frame loop only once it is past the last window. Count in on the song's own beat grid. [references/timing.md](references/timing.md) has the code.
69
+ - **Reference configuration:** OH snaps the clock when it is more than 40 ms off and otherwise glides at 0.08 a frame; lead-in 2.6 s; resume count-in `max(1.2, min(2.4, 3 beats))` s. Real key presses scheduled off the audio clock landed at 1.1 ms mean and 2.5 ms p95 with no misses, 1.8 ms mean under a 250 ms stall every 1.5 s, and 2.1 ms at 4x CPU throttle; drawn and heard stayed within 4 ms. Every composed song's first beat sat at 0.30 s while the count-in counted from 0, about 62% of a beat off at 124 bpm, until QA swept every song.
70
+ - **Applies when:** Any game that judges input against music.
71
+ - **Verify:** Real key presses timed off the audio clock, not self-play (which lands every note at zero): mean under 3 ms, p95 under 5 ms, no misses under stalls or CPU throttling.
72
+
73
+ ### Hit windows are a player setting; difficulty is density and travel
74
+ - **Invariant:** The timing windows are the same on every level and set by the player; levels differ in density and travel time; a note-speed setting changes only how fast notes travel.
75
+ - **Technique:** Offer relaxed, normal and strict windows. Set travel time per level and divide it by the note-speed setting. Keep a separate picture offset that moves only the drawn notes.
76
+ - **Reference configuration:** OH normal windows are ±46, 90 and 138 ms (relaxed 62, 122 and 185; strict 32, 62 and 95), against a genre reference of about 116 ms early and 100 late. Travel is 2.6, 2.3, 2.05 and 1.85 s by level, note speed 0.5 to 2.5x; the genre scrolls 3.5 to 2.5 s. Accuracy is `(P + 0.7G + 0.4Gd) / judged`, and S+ is a full combo above 0.97.
77
+ - **Applies when:** Timing-judged games with levels.
78
+ - **Verify:** A unit test that the windows follow the setting and not the level, and a timed run at each window setting.
79
+
80
+ ### Calibrate by sight and sound, and never as an answer to a complaint
81
+ - **Invariant:** Latency is set by lining up a visible landing with a heard click, offered before the first song, in plain words; a complaint about the charts is never answered with "calibrate first".
82
+ - **Technique:** A dot falls onto a ring with a click scheduled on the play clock; arrow keys nudge it; an in-play nudge key is there for later. Both the ring and the dot must be fully drawn at the landing frame.
83
+ - **Reference configuration:** OH replaced a 12-tap test after eight hours because "a tap test measures the tapper as much as the speakers", and because the owner wrote "i didnt run calibrate idk what that even means". The sound check moves 10 ms per left or right, 1 ms per up or down, over -200 to +400 ms, with `[` and `]` nudging 10 ms in play and a picture offset of -100 to +100 ms. Its first version failed because the ring faded in at the moment of landing.
84
+ - **Applies when:** Every rhythm game.
85
+ - **Verify:** Offsets of ±80 ms keep every timed press Perfect, and a still of the landing frame shows both ring and dot.
86
+
87
+ ### The board is sacred
88
+ - **Invariant:** Notes, catchers and the strike line draw every frame, sharp, with no lens or post effect on them, from the count-in on; no world effect covers a lane; lane contrast against the board is measured, not judged.
89
+ - **Technique:** Draw the board to its own full-resolution target and composite it last over the world. Mask every world effect out of a corridor round the lanes. Keep the HUD quieter than the notes. Draw one composite to the canvas per frame (`opengame-stability`).
90
+ - **Reference configuration:** OH's board fills 35% of the picture height, far notes at 42% of near size; lane colours `#ff5436`, `#ffb41f`, `#1fc9a2`, `#2f86ff`, `#f04cc8` at L* 56.8 to 79.3, each at least 28 L* above the board's lightest tenth (a lighter board once cut blue to 2.6:1). Two draws to the screen a frame lost the board for single frames, 4 of 880; one draw lost 0 of about 7,600.
91
+ - **Applies when:** Any lane game in front of a busy or stylised world.
92
+ - **Verify:** An in-page probe at a catcher over two minutes in two songs shows no missing frame, and an L* table per venue.
93
+
94
+ ### Notes read clearly on a receding highway
95
+ - **Invariant:** Every note variant shares one head width and timing centre; marks lying on the board stay thin, dark and along the lane; notes in a dense stream never touch; a hold reads as a ribbon, not a tube.
96
+ - **Technique:** Vary paint inside fixed bounds, and cap each note's length by its gap to the nearest neighbour, counting a neighbour in the next lane as further away.
97
+ - **Reference configuration:** OH: on a receding board "anything bright or tall stands up" (a bright tongue read as a fin, streaks as flames, dark tapers as teeth). Tails a tenth of the head wide and at most two heads long worked; holds are a flat ribbon 60% of the lane ending in an inked bar ("a round pale end read as a lid on a tube"); the next-lane neighbour counts 1.6x further; tested at 9 notes a second.
98
+ - **Applies when:** Designing notes, tails, holds and hit splats.
99
+ - **Verify:** 2x crops at the strike line and at the far end of the fastest passage in the catalogue.
100
+
101
+ ### Scoring rewards streaks, and star power is a board-wide state
102
+ - **Invariant:** A streak multiplier is the second thing a player reads after the notes; star power changes the whole board while notes keep their lane colours; its charge, its ready state with its key, and its time left are always readable; failing is optional.
103
+ - **Technique:** Charge star power from clean phrases and let it fire from half full for double points. Show charge, then READY with the key until used, then only the reward and a countdown. Words leave the screen by shrinking and drifting while opaque, never by ghosting.
104
+ - **Reference configuration:** OH: Perfect, Great and Good score 100, 70 and 40 times `min(4, 1 + floor(combo / 10))`, doubled in star power; a clean phrase adds 0.25 of charge; drain `dt / max(8, 32 * 60 / bpm)`; holds pay 60 a second of song time. READY never showed under autoplay because self-play fired it at once; a real-key player first saw it at 31.36 s. A miss dulls the music with a lowpass to 520 Hz in 30 ms, back over 340 ms.
105
+ - **Applies when:** Any streak, combo or power mechanic.
106
+ - **Verify:** A real-key run shows READY until pressed, and a self-play bot holds its charge two seconds before firing so captures can see it.
107
+
108
+ ### The world moves with the song's structure, eased, never shaken
109
+ - **Invariant:** The world changes on bars and four-bar phrases, the camera never shakes, the board never moves, every hit answers in its own lane at once, and a full-screen hit moment is rare and timed by the song clock.
110
+ - **Technique:** Put one music clock (beat phase, energy, bands) in shared shader uniforms. Make one camera move every four bars and a small downbeat push, applied after any framing solve so the solve cannot undo it. Use asymmetric moving parts in open sky so motion reads. Decide big hit moments in one tested function that requesters only ask. No hitstop in a timing game.
111
+ - **Reference configuration:** OH: downbeat push 2.5% (3% on a phrase line) in 0.25 s and out over 1.35 s; a 3 degree pan per phrase; a palette wipe 1.6 s per phrase. Per hit: splat 0.12 s, halo 0.1 s, a tick of band-pass noise at `2200 + 420 * lane` Hz for 35 ms. The hit frame is ink 70 ms, flat 70 ms, wipe 120 ms, at most once per 8 beats and 6 s, after a run of 8. The owner on background motion: "the background is hsaking so much ... just looks like slop"; judges had been recommending a 6 px shake until the owner's rule reached their rubric. The framing solve silently undid the camera life three times.
112
+ - **Applies when:** Any music-synced scene.
113
+ - **Verify:** 1:1 strips across a bar line, a strike log over forty seconds of play, and the calm setting probed for stillness.
114
+
115
+ ### The charted attack is heard in the final mix, on time
116
+ - **Invariant:** Every charted note's attack is audible over the whole mix at its charted time, and the decoded file the game plays carries no codec padding into the song clock.
117
+ - **Technique:** Trim samples to onset, duck backing and reverb under the lead, lift featured onsets, and double a soft lead with a percussive voice a few ms ahead. Master after encoding is checked, and shift for encoder delay only when the tag and the decoded audio agree.
118
+ - **Reference configuration:** OH gates: featured stem onset p95 under 10 ms; whole mix p90 under 30 ms with a 50 ms search; attack margin median 3 dB with 90% above zero; onset lift 2.5 to 6 dB; -14 LUFS ±0.6 at most -0.99 dBTP measured on the decoded MP3; 256 kbit/s CBR; 25.06 ms of codec silence trimmed only when every explanation agreed within 0.3 ms.
119
+ - **Applies when:** Games that judge a player against audio.
120
+ - **Verify:** A mix probe with each stem left out in turn, and clicks placed at known times decoded in every browser you can run, within 2 ms; name the ones you could not.
121
+
122
+ ### Check every chart through the game's own code
123
+ - **Invariant:** The tools, the judges and the player all see the same chart, note for note, and a missing or broken chart is an error the player reads, never a guessed replacement.
124
+ - **Technique:** Bundle the game's chart code for Node and run every song at every level in seconds: bad times and lanes, overlaps, holds into the next note, fast repeats, density by level, length against the audio. Compare the results screen's note count with the chart file.
125
+ - **Reference configuration:** OH checked 169 songs at 4 levels in about 2 s. A hidden in-game filter removed 20 to 26% of notes on quiet and fast songs for 7 h 20 min while judged charts were complete; two judges caught it from a results screen showing 117 notes against 147 in the grid. A missing chart file once played 717 guessed notes in place of 403.
126
+ - **Applies when:** Any game with chart or level data.
127
+ - **Verify:** The game-reported note count equals the file's for every song QA plays to the end, and an injected missing file shows a message.
128
+
129
+ ## The owner is the only ear
130
+
131
+ No critic, benchmark or self-play run in OH heard a song or pressed a key. Only the owner
132
+ found the chart quality, impossible Expert, the background shake, the boring songs and
133
+ everything about sound. Every update gives him one link, one named song and the hardest
134
+ level, and never reports a proxy score as progress. The loop is in `opengame-production`.