@opengamehq/skills 1.4.0 → 1.6.0

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 (383) hide show
  1. package/CHANGELOG.md +36 -0
  2. package/README.md +12 -4
  3. package/dist/.claude-plugin/marketplace.json +26 -0
  4. package/dist/skills/manifest.json +1407 -143
  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 +25 -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 +108 -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 +190 -0
  25. package/dist/skills/opengame-all-in-one/references/opengame-deploy/README.md +229 -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 +90 -0
  28. package/dist/skills/opengame-all-in-one/references/opengame-deploy/references/packaging.md +120 -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 +121 -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 +12 -0
  185. package/dist/skills/opengame-api/SKILL.md +40 -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 +16 -0
  202. package/dist/skills/opengame-cli/SKILL.md +13 -11
  203. package/dist/skills/opengame-deploy/CHANGELOG.md +24 -0
  204. package/dist/skills/opengame-deploy/SKILL.md +47 -21
  205. package/dist/skills/opengame-deploy/references/json-events.md +6 -5
  206. package/dist/skills/opengame-deploy/references/manifest.md +3 -2
  207. package/dist/skills/opengame-deploy/references/packaging.md +29 -26
  208. package/dist/skills/opengame-depth/CHANGELOG.md +8 -0
  209. package/dist/skills/opengame-depth/SKILL.md +61 -33
  210. package/dist/skills/opengame-enemies/CHANGELOG.md +10 -0
  211. package/dist/skills/opengame-enemies/SKILL.md +96 -40
  212. package/dist/skills/opengame-enemies/references/pacing-and-bots.md +108 -0
  213. package/dist/skills/opengame-enemies/references/threejs.md +59 -0
  214. package/dist/skills/opengame-front-end/CHANGELOG.md +4 -0
  215. package/dist/skills/opengame-front-end/SKILL.md +85 -39
  216. package/dist/skills/opengame-front-end/references/menus.md +42 -8
  217. package/dist/skills/opengame-game-feel/CHANGELOG.md +11 -0
  218. package/dist/skills/opengame-game-feel/SKILL.md +95 -47
  219. package/dist/skills/opengame-game-feel/references/feel-numbers.md +59 -0
  220. package/dist/skills/opengame-game-feel/references/killcam.md +88 -0
  221. package/dist/skills/opengame-jam/CHANGELOG.md +12 -0
  222. package/dist/skills/opengame-jam/SKILL.md +59 -37
  223. package/dist/skills/opengame-lighting/CHANGELOG.md +9 -0
  224. package/dist/skills/opengame-lighting/SKILL.md +85 -34
  225. package/dist/skills/opengame-loading/CHANGELOG.md +4 -0
  226. package/dist/skills/opengame-loading/SKILL.md +78 -38
  227. package/dist/skills/opengame-loading/helpers/loading-count.mjs +115 -0
  228. package/dist/skills/opengame-loading/references/threejs.md +32 -7
  229. package/dist/skills/opengame-look/CHANGELOG.md +9 -0
  230. package/dist/skills/opengame-look/SKILL.md +122 -51
  231. package/dist/skills/opengame-look/references/matching-a-reference.md +73 -0
  232. package/dist/skills/opengame-look/references/painterly.md +69 -0
  233. package/dist/skills/opengame-look/references/post-and-lens.md +10 -3
  234. package/dist/skills/opengame-maps/CHANGELOG.md +9 -0
  235. package/dist/skills/opengame-maps/SKILL.md +78 -0
  236. package/dist/skills/opengame-maps/references/map-spec.md +77 -0
  237. package/dist/skills/opengame-mobile/CHANGELOG.md +4 -0
  238. package/dist/skills/opengame-mobile/SKILL.md +18 -18
  239. package/dist/skills/opengame-montage-make/CHANGELOG.md +5 -0
  240. package/dist/skills/opengame-montage-make/SKILL.md +187 -0
  241. package/dist/skills/opengame-montage-make/references/briefs.md +293 -0
  242. package/dist/skills/opengame-montage-make/references/day-one-questions.md +46 -0
  243. package/dist/skills/opengame-montage-make/references/edge-cases.md +27 -0
  244. package/dist/skills/opengame-montage-make/references/filmable.md +244 -0
  245. package/dist/skills/opengame-montage-make/references/finishing.md +126 -0
  246. package/dist/skills/opengame-montage-make/references/measure.md +282 -0
  247. package/dist/skills/opengame-montage-make/references/montage-capture.md +151 -0
  248. package/dist/skills/opengame-montage-make/references/montage-gates.md +84 -0
  249. package/dist/skills/opengame-montage-make/references/montage-genres.md +198 -0
  250. package/dist/skills/opengame-montage-make/references/montage-grammar.md +320 -0
  251. package/dist/skills/opengame-montage-make/references/montage-ops.md +127 -0
  252. package/dist/skills/opengame-montage-make/references/og-snipers-record.md +110 -0
  253. package/dist/skills/opengame-montage-make/references/owner-eye.md +214 -0
  254. package/dist/skills/opengame-montage-make/references/reference-study.md +293 -0
  255. package/dist/skills/opengame-montage-make/references/song.md +64 -0
  256. package/dist/skills/opengame-montage-plan/CHANGELOG.md +5 -0
  257. package/dist/skills/opengame-montage-plan/SKILL.md +197 -0
  258. package/dist/skills/opengame-montage-plan/references/briefs.md +293 -0
  259. package/dist/skills/opengame-montage-plan/references/day-one-questions.md +46 -0
  260. package/dist/skills/opengame-montage-plan/references/filmable.md +244 -0
  261. package/dist/skills/opengame-montage-plan/references/measure.md +282 -0
  262. package/dist/skills/opengame-montage-plan/references/montage-capture.md +151 -0
  263. package/dist/skills/opengame-montage-plan/references/montage-gates.md +84 -0
  264. package/dist/skills/opengame-montage-plan/references/montage-genres.md +198 -0
  265. package/dist/skills/opengame-montage-plan/references/montage-grammar.md +320 -0
  266. package/dist/skills/opengame-montage-plan/references/montage-ops.md +127 -0
  267. package/dist/skills/opengame-montage-plan/references/og-snipers-record.md +110 -0
  268. package/dist/skills/opengame-montage-plan/references/owner-eye.md +214 -0
  269. package/dist/skills/opengame-montage-plan/references/reference-study.md +293 -0
  270. package/dist/skills/opengame-montage-plan/references/song.md +64 -0
  271. package/dist/skills/opengame-montage-review-and-improve/CHANGELOG.md +5 -0
  272. package/dist/skills/opengame-montage-review-and-improve/SKILL.md +178 -0
  273. package/dist/skills/opengame-montage-review-and-improve/references/briefs.md +293 -0
  274. package/dist/skills/opengame-montage-review-and-improve/references/day-one-questions.md +46 -0
  275. package/dist/skills/opengame-montage-review-and-improve/references/filmable.md +244 -0
  276. package/dist/skills/opengame-montage-review-and-improve/references/finishing.md +126 -0
  277. package/dist/skills/opengame-montage-review-and-improve/references/measure.md +282 -0
  278. package/dist/skills/opengame-montage-review-and-improve/references/montage-capture.md +151 -0
  279. package/dist/skills/opengame-montage-review-and-improve/references/montage-gates.md +84 -0
  280. package/dist/skills/opengame-montage-review-and-improve/references/montage-genres.md +198 -0
  281. package/dist/skills/opengame-montage-review-and-improve/references/montage-grammar.md +320 -0
  282. package/dist/skills/opengame-montage-review-and-improve/references/montage-ops.md +127 -0
  283. package/dist/skills/opengame-montage-review-and-improve/references/og-snipers-record.md +110 -0
  284. package/dist/skills/opengame-montage-review-and-improve/references/owner-eye.md +214 -0
  285. package/dist/skills/opengame-montage-review-and-improve/references/reference-study.md +293 -0
  286. package/dist/skills/opengame-montage-review-and-improve/references/song.md +64 -0
  287. package/dist/skills/opengame-multiplayer/CHANGELOG.md +15 -0
  288. package/dist/skills/opengame-multiplayer/SKILL.md +143 -0
  289. package/dist/skills/opengame-multiplayer/references/bot-lobbies.md +71 -0
  290. package/dist/skills/opengame-multiplayer/references/netcode.md +165 -0
  291. package/dist/skills/opengame-performance/CHANGELOG.md +8 -0
  292. package/dist/skills/opengame-performance/SKILL.md +106 -65
  293. package/dist/skills/opengame-performance/helpers/landing-pacer.mjs +86 -0
  294. package/dist/skills/opengame-performance/references/measuring.md +51 -0
  295. package/dist/skills/opengame-performance/references/webgpu.md +86 -0
  296. package/dist/skills/opengame-production/CHANGELOG.md +13 -0
  297. package/dist/skills/opengame-production/SKILL.md +119 -54
  298. package/dist/skills/opengame-production/references/team.md +107 -0
  299. package/dist/skills/opengame-production/references/templates.md +106 -16
  300. package/dist/skills/opengame-quality-bar/CHANGELOG.md +10 -0
  301. package/dist/skills/opengame-quality-bar/SKILL.md +94 -42
  302. package/dist/skills/opengame-quality-bar/references/cheap-tells.md +14 -7
  303. package/dist/skills/opengame-quality-bar/references/judges.md +95 -0
  304. package/dist/skills/opengame-quality-bar/references/review-prompt.md +10 -5
  305. package/dist/skills/opengame-rhythm/CHANGELOG.md +5 -0
  306. package/dist/skills/opengame-rhythm/SKILL.md +138 -0
  307. package/dist/skills/opengame-rhythm/references/charting.md +109 -0
  308. package/dist/skills/opengame-rhythm/references/timing.md +92 -0
  309. package/dist/skills/opengame-router/CHANGELOG.md +8 -0
  310. package/dist/skills/opengame-router/SKILL.md +14 -7
  311. package/dist/skills/opengame-scene-audit/CHANGELOG.md +15 -0
  312. package/dist/skills/opengame-scene-audit/SKILL.md +64 -47
  313. package/dist/skills/opengame-scene-audit/helpers/scene-audit.js +97 -19
  314. package/dist/skills/opengame-sdk/CHANGELOG.md +12 -0
  315. package/dist/skills/opengame-sdk/SKILL.md +26 -9
  316. package/dist/skills/opengame-sdk/references/engines.md +2 -2
  317. package/dist/skills/opengame-sdk/references/identity.md +10 -11
  318. package/dist/skills/opengame-stability/CHANGELOG.md +4 -0
  319. package/dist/skills/opengame-stability/SKILL.md +59 -27
  320. package/dist/skills/opengame-trailer-make/CHANGELOG.md +5 -0
  321. package/dist/skills/opengame-trailer-make/SKILL.md +164 -0
  322. package/dist/skills/opengame-trailer-make/references/archetypes.md +113 -0
  323. package/dist/skills/opengame-trailer-make/references/briefs.md +273 -0
  324. package/dist/skills/opengame-trailer-make/references/core.md +129 -0
  325. package/dist/skills/opengame-trailer-make/references/craft.md +299 -0
  326. package/dist/skills/opengame-trailer-make/references/edge-cases.md +27 -0
  327. package/dist/skills/opengame-trailer-make/references/filmable.md +244 -0
  328. package/dist/skills/opengame-trailer-make/references/finishing.md +126 -0
  329. package/dist/skills/opengame-trailer-make/references/gates.md +184 -0
  330. package/dist/skills/opengame-trailer-make/references/measure.md +282 -0
  331. package/dist/skills/opengame-trailer-make/references/music.md +148 -0
  332. package/dist/skills/opengame-trailer-make/references/og-clash-record.md +133 -0
  333. package/dist/skills/opengame-trailer-make/references/ops.md +154 -0
  334. package/dist/skills/opengame-trailer-make/references/owner-eye.md +214 -0
  335. package/dist/skills/opengame-trailer-make/references/review.md +141 -0
  336. package/dist/skills/opengame-trailer-make/references/title.md +92 -0
  337. package/dist/skills/opengame-trailer-plan/CHANGELOG.md +5 -0
  338. package/dist/skills/opengame-trailer-plan/SKILL.md +176 -0
  339. package/dist/skills/opengame-trailer-plan/references/archetypes.md +113 -0
  340. package/dist/skills/opengame-trailer-plan/references/briefs.md +273 -0
  341. package/dist/skills/opengame-trailer-plan/references/core.md +129 -0
  342. package/dist/skills/opengame-trailer-plan/references/craft.md +299 -0
  343. package/dist/skills/opengame-trailer-plan/references/edge-cases.md +27 -0
  344. package/dist/skills/opengame-trailer-plan/references/filmable.md +244 -0
  345. package/dist/skills/opengame-trailer-plan/references/finishing.md +126 -0
  346. package/dist/skills/opengame-trailer-plan/references/gates.md +184 -0
  347. package/dist/skills/opengame-trailer-plan/references/measure.md +282 -0
  348. package/dist/skills/opengame-trailer-plan/references/music.md +148 -0
  349. package/dist/skills/opengame-trailer-plan/references/og-clash-record.md +133 -0
  350. package/dist/skills/opengame-trailer-plan/references/ops.md +154 -0
  351. package/dist/skills/opengame-trailer-plan/references/owner-eye.md +214 -0
  352. package/dist/skills/opengame-trailer-plan/references/reference-set.md +92 -0
  353. package/dist/skills/opengame-trailer-plan/references/review.md +141 -0
  354. package/dist/skills/opengame-trailer-plan/references/title.md +92 -0
  355. package/dist/skills/opengame-trailer-review-and-improve/CHANGELOG.md +5 -0
  356. package/dist/skills/opengame-trailer-review-and-improve/SKILL.md +145 -0
  357. package/dist/skills/opengame-trailer-review-and-improve/references/archetypes.md +113 -0
  358. package/dist/skills/opengame-trailer-review-and-improve/references/briefs.md +273 -0
  359. package/dist/skills/opengame-trailer-review-and-improve/references/core.md +129 -0
  360. package/dist/skills/opengame-trailer-review-and-improve/references/craft.md +299 -0
  361. package/dist/skills/opengame-trailer-review-and-improve/references/edge-cases.md +27 -0
  362. package/dist/skills/opengame-trailer-review-and-improve/references/filmable.md +244 -0
  363. package/dist/skills/opengame-trailer-review-and-improve/references/finishing.md +126 -0
  364. package/dist/skills/opengame-trailer-review-and-improve/references/gates.md +184 -0
  365. package/dist/skills/opengame-trailer-review-and-improve/references/measure.md +282 -0
  366. package/dist/skills/opengame-trailer-review-and-improve/references/music.md +148 -0
  367. package/dist/skills/opengame-trailer-review-and-improve/references/og-clash-record.md +133 -0
  368. package/dist/skills/opengame-trailer-review-and-improve/references/ops.md +154 -0
  369. package/dist/skills/opengame-trailer-review-and-improve/references/owner-eye.md +214 -0
  370. package/dist/skills/opengame-trailer-review-and-improve/references/review.md +141 -0
  371. package/dist/skills/opengame-trailer-review-and-improve/references/title.md +92 -0
  372. package/dist/skills/opengame-verification/CHANGELOG.md +11 -0
  373. package/dist/skills/opengame-verification/SKILL.md +113 -48
  374. package/dist/skills/opengame-weapons/CHANGELOG.md +11 -0
  375. package/dist/skills/opengame-weapons/SKILL.md +129 -55
  376. package/dist/skills/opengame-weapons/references/handling-numbers.md +42 -0
  377. package/dist/skills/opengame-weapons/references/reloads-and-hands.md +96 -0
  378. package/dist/skills/opengame-weapons/references/sights-and-scopes.md +98 -0
  379. package/dist/skills/opengame-weapons/references/threejs.md +71 -0
  380. package/dist/skills/opengame-world-physics/CHANGELOG.md +9 -0
  381. package/dist/skills/opengame-world-physics/SKILL.md +55 -24
  382. package/dist/skills/opengame-world-physics/references/measured-failures.md +28 -1
  383. package/package.json +1 -1
@@ -0,0 +1,293 @@
1
+ # Studying the reference edit: the golden edit, its finishing layer, and a good player
2
+
3
+ The owner usually gives one reference edit ("edit it like this") and treats it as the bar: 10/10 by definition. Measure
4
+ it before planning, as a study a judge can score against. The worked example is the OG Snipers study of Kaiser's FaZe
5
+ sniping edit "FAINT" (95.24 s, 2,858 frames at 1080p30, 4 Mbps). Download it privately for measurement only; never cut
6
+ it into your film, play it aloud, or recreate its game, characters, logo or footage.
7
+
8
+ **Units.** Its frame numbers and frame counts are at 30 fps: double them for a 60 fps timeline. Its speeds in px/f and
9
+ its angles per frame are per 30 fps frame at 1080p: halve them for a 60 fps timeline (FAINT's incoming seams at a median
10
+ 27 px/f are about 13.5 px/f at 60 fps).
11
+
12
+ ## 1. The shape of a reference study
13
+
14
+ Write `research/REF-STUDY.md` with these sections, every number measured from the file (the measuring chapter) or
15
+ marked "by eye":
16
+
17
+ - **A. Identity:** the file, its length, rate, bitrate; its structure by song section.
18
+ - **B. A shot log:** every shot with frames, type (first-person gameplay, a cinematic tied to a clip, a free cinematic,
19
+ title or text), what happens, and the transition in.
20
+ - **C. Measured statistics:** cuts and shot lengths; the screen-time split; the gameplay clips; sync to the music; speed
21
+ ramps; grade; letterbox; sound.
22
+ - **D. The cinematics in detail:** the cold open; the cinematics that replay or extend a clip; the free ones.
23
+ - **E. A transitions catalogue:** every kind, its count, its frames, what in the picture causes it.
24
+ - **F. Text and titles.**
25
+ - **G. Flow and aura as rules with numbers.**
26
+ - **H. A rubric for the judges,** 0-10 per dimension, with what 10, 7 and 5 look like in this file.
27
+ - **I. A steal list:** 20-30 techniques with frames and how to reproduce each in an engine that owns the camera and the
28
+ clock.
29
+ - **What could not be measured,** said in so many words.
30
+
31
+ Check automatic cut detection against every frame by eye: on FAINT the detector returned 75 cuts, of which 20 were flashes
32
+ inside a shot, and missed about 25 real cuts hidden in whips, irises, portals and white frames. The corrected list had
33
+ 68 real cuts and 22 transition-only inserts.
34
+
35
+ ## 2. FAINT, measured
36
+
37
+ **Cuts and shots.** 68 cuts in 95.27 s, 0.71 a second; nearly flat after the cold open (4-9 per 10 s), because the
38
+ energy comes from what is inside the shots. Shot lengths: mean 1.35 s, median 1.03 s; 23 % under 0.5 s, 49 % under 1 s,
39
+ 19 % over 2 s. First-person gameplay segments: median 1.37 s, up to 5.17 s without a cut. Cinematics tied to a clip:
40
+ median 0.50 s. The rhythm inside gameplay is whips, not cuts: optical flow peaks 175 times (1.84 a second), each rising
41
+ in 3 frames and falling in 3. No held or frozen frames: 2 near-duplicate pairs in 2,857 (0.07 %).
42
+
43
+ **The screen-time split** (what the owner asked OG Snipers to follow): over the whole file 52 % first-person gameplay, 37 %
44
+ cinematic (12 % tied to a clip, 26 % free), 9 % titles and text, 2 % transitions. It is front-loaded: the first 32 s hold
45
+ 12.3 s of story cinematic, 8 s of titles and 5.2 s of gameplay. After the titles the body runs 70 % first person, 24 %
46
+ cinematic, 3 % text, 3 % transitions. The file is cut off, so an outro would raise the cinematic share.
47
+
48
+ **The clips.** 11 clips; first-person time per clip a median 4.87 s (1.40-7.13); kills per clip a median 5 (3-9 in the
49
+ ten complete clips); one kill per 0.95 s of first-person screen time. Kill spacing inside a clip: 2 beats is the mode (18
50
+ of 46 gaps), then 1.5 beats (8), 4 (5), 2.5 and 3 (4 each), 1 (3). Quickscopes are the default (about 40 of 52); each clip
51
+ has one twist (a wall-jump shot, a pistol finisher, a match-winning kill, a held scope, a collateral shown by a bullet
52
+ camera, a slide, backflip and spin into a charged shot). Inside a first-person segment the 95th-percentile flow is a
53
+ median 13.8x the 10th: every clip swings between near-still holds and very fast whips.
54
+
55
+ **Sync.** One constant 135.0 BPM grid (a beat is 13.33 frames at 30 fps, 26.67 at 60). As delivered almost nothing sits
56
+ on a beat's own frame; shift the picture 2 frames later and 70 % of shot frames and 40 % of cuts sit within a frame of a
57
+ beat. **The picture leads the beat by 1.5-2 reference frames: 3-4 frames at 60 fps.** 51 % of cuts fall in the last
58
+ quarter of a beat (25 % would be even). Cut spacing varies (1 to 12 beats; the edit never cuts on a fixed count). Motion
59
+ is beat-locked: folded onto the beat, flow is highest in the last 30 % before it, a swing of 2.49x over the film and 3.23x
60
+ inside gameplay. Every section change in the music is answered in the picture 2-3 frames early (the drop by ink frames
61
+ and a light-streak zoom, the chorus by a flash, the second verse by a whip-cut). Flashes: 45 luma jumps over 35 levels,
62
+ 28.3 a minute, a median half-decay of 3 frames.
63
+
64
+ **Ramps.** Around a shot (flow relative to the film's 90th percentile, frame 0 the shot, at 30 fps): −10 0.28, −8 0.23,
65
+ −6 0.18, −4 0.28, −3 0.35, −2 0.63, −1 0.94, **0 0.84**, +1 0.88, +2 0.70, +3 0.79, +4 0.54, +5 0.56, +6 0.43, +7 0.31,
66
+ +8 0.32. A glide at 0.18-0.28 until 4 frames before; a rise over 3 frames; the peak held 5 frames (flick, shot, recoil); a
67
+ fall over 4. Peak over glide about 5x. At 60 fps: rise 6, peak 10, fall 8. Around a cut, the outgoing shot is calm and
68
+ lifts only in its last frame or two, and **the incoming shot arrives at about twice that speed and settles over 5
69
+ frames**: shots are entered fast and left slow. Slow holds in gameplay: a median 11 frames (0.37 s), the longest 21 (the
70
+ airborne hold on the drop). Long near-still spans exist only in the cold open and titles.
71
+
72
+ **Grade.** One look for the whole film: a teal cast (a* about −12 in every section), warm things left warm so they pop.
73
+ Two brightness worlds: gameplay dark and contrasty (luma 98, 10 % crushed), cinematics high-key (luma 125-158). **The
74
+ median luma jump across a cut is 37 levels, and 29 of 68 cuts jump more than 40:** the edit alternates dark and bright on
75
+ purpose. That breaks the trailer gate of a median jump at or under 18; for this style, waive the gate at
76
+ gameplay-to-cinematic cuts once the owner has approved the look. Almost nothing clips outside flash frames.
77
+
78
+ **Sound.** The song runs unbroken at one tempo, with one big contrast: 16 s of quiet intro, then the drop. Gunshots are
79
+ not louder than the music's own hits (onset strength at the shot frames sits below that of the beats), which is the
80
+ owner's "the music leads" in numbers. One spoken line, a character's, in the cold open.
81
+
82
+ **Letterbox.** 2.20:1 bars on the cold open only, opening over 28 frames as the title lands. Decide it on a strip of your
83
+ own cold open; never assume it.
84
+
85
+ **Endings.** The attached file is cut off and has no ending. The other Kaiser edits measured for OG Clash all end on a
86
+ held, slow image (0-0.3 cuts a second): a tableau, black and silence, then a calm title.
87
+
88
+ ## 3. Flow and aura, as rules
89
+
90
+ 1. Speed is carried across the cut; stillness never meets stillness (incoming shots start at a median 27 px/f at 30 fps;
91
+ only 3 of 68 seams are still on both sides, all in the cold open or titles). Direction need not match (a median 115°
92
+ apart).
93
+ 2. Arrive fast, settle in 5 frames.
94
+ 3. Every kill breathes the same way: glide, rise 3, peak 5, fall 4. The repetition of one shape reads as smooth.
95
+ 4. Motion lives in the last third of the beat.
96
+ 5. The picture leads the sound by 2 frames at 30 fps.
97
+ 6. Nothing ever stops: no held frames, no freeze frames; the slowest spans still drift or sway.
98
+ 7. Density alternates by song section in content, not cut rate: verses (rapped) are 77-94 % gameplay at nearly a kill a
99
+ second; choruses (sung, 5 LU louder) are twice as cinematic with the fastest cutting and the set pieces. The loudest
100
+ music gets the most cinema, not the most kills, and the loudest change in the music meets the least motion (the drop
101
+ lands on an airborne slow hold).
102
+ 8. Gameplay runs 3-7 s, then a short punctuation of 0.4-2.9 s; ten of eleven clips have a cinematic attached.
103
+ 9. One real breath, late: 5.2 s without gameplay after 16 s that were 94 % gameplay, before the last lift.
104
+ 10. No metronome: the longest run of equally spaced cuts is 3.
105
+ 11. Dark and bright alternate: a return to gameplay feels like a hit because the picture drops into contrast.
106
+ 12. One colour, one hero: the same cast in every 10 s window, the same character in every cinematic.
107
+ 13. Effects are brief and each kind is rationed: flashes decay in 3 frames; no transition type except the whip appears
108
+ more than 7 times; signature devices (a glass-shard portal, a bullet camera) once each.
109
+ 14. Every transition is caused by something in the picture: a muzzle flash, a scope, a lens, a shard, a sleeve, a boot.
110
+ 15. Titles and effects never cover a kill: tint flashes start on the kill frame, not before.
111
+ 16. Two lens families: ultra-wide, low, tilted and inches away for action; long and shallow for calm moments and the
112
+ weapon. The subject is off the ground in action shots. Something new every 6-10 s.
113
+
114
+ **Where you can beat it:** it is 30 fps and 4 Mbps; its whole HUD is on; its master clips (true peak +0.7 dBFS); one teal
115
+ wash over eleven maps makes places hard to tell apart and turns skin olive; 12.3 s of titles hold back the gameplay;
116
+ three dead seams; a few dated devices (two-frame negative flashes, a text decode). Its music never breathes after the
117
+ drop; a trailer would add a designed silence there, but a montage on the owner's song must not ("u cant have music just
118
+ go completely blank"). What cannot be copied: its game's characters, animations and set pieces, a real player's recorded
119
+ matches, its track and its emblem.
120
+
121
+ ## 4. The rubric anchors (what 10, 7 and 5 look like)
122
+
123
+ | Dimension | 10 in FAINT | 7 | 5 |
124
+ |---|---|---|---|
125
+ | Clip quality | multi-kills by one visibly human player: a median of 5, a kill per 0.95 s, a twist per clip, no miss shown | 3-5 kills, clean, the same pattern every clip | 1-3 kills, or kills that look scripted |
126
+ | Music sync | 70 % of shot frames within a frame of a beat at the lead; section changes answered within 3 frames; swing 3.2x in gameplay | 45-60 % on a beat; swing about 2x | under 35 %; swing under 1.5x |
127
+ | Speed ramps | one repeated shape, peak over glide about 5x, holds of 0.37 s, no doubled frames | smooth but 2-3x contrast, or holds of a second that read as pauses | constant speed, interpolation smear, freezes |
128
+ | Transitions | 16 kinds, each caused by the picture; 3 dead seams of 68 | 6-8 kinds, mostly whips and flashes; one signature piece; under 10 % dead seams | hard cuts plus stock overlays; 20 % or more dead seams; dissolves |
129
+ | Cinematics | 37 % of the file; 19 shots replay a clip at a median 0.50 s; kills from the victim's side | good-looking but separate from the clips | orbits of a standing character; no replay of any kill |
130
+ | Camera | the two lens families; a one-take reveal with a foreground wipe | varied, some low and wide, several eye-level | one lens, eye level, moves that start or end on a stop |
131
+ | Grade | one world, dark gameplay against high-key cinematics, under 0.1 % clipped | consistent, but no contrast between gameplay and cinematics | ungraded, or a different look per shot |
132
+ | Pacing | cold open, a drop on an airborne hold, verses of gameplay, a cinematic chorus, a late breath | sections exist but the calm parts are short or missing | one density throughout |
133
+ | Flow | nothing stops; arrive fast, settle in 5; the same ease on every kill | a few abrupt stops or a metronomic stretch | snaps without easing, dead frames at cuts |
134
+ | Overall aura | one authored piece: one colour, one hero, one ramp shape, the song visible in the picture | strong with a flat stretch | a compilation |
135
+
136
+ ## 5. A steal list from FAINT (reproduced with an engine that owns the camera and the clock)
137
+
138
+ Frame counts here are already at 60 fps. "Sim speed" is simulation time per output frame, stepped by the engine, never
139
+ interpolated.
140
+
141
+ 1. **The kill cycle:** sim speed 0.2 until 8 frames before the shot, ease to 1.0 over 6, hold 1.0 from 2 before to 6 after
142
+ (flick, shot, recoil), ease back over 8. Check the peak's median flow is at least 4x the glide's.
143
+ 2. **Kills two beats apart, the shot frame 4 frames before the beat:** solve the clock so the shots land there; never
144
+ slide the audio.
145
+ 3. **A scope-shot exposure bump:** luma up about 18 of 255 on the shot frame, settling over 16 frames, tied to the muzzle
146
+ light.
147
+ 4. **Muzzle flash, one white frame, first person:** end the outside shot on its muzzle frame, one frame near white,
148
+ start first person already moving.
149
+ 5. **First person to third person in the air and back,** in one simulation take: at the jump's apex switch to a camera
150
+ 0.5-1 m from the body, 14-18 mm, low and tilted; sim speed 0.2 for about 24 frames on the aim; cut back while still
151
+ airborne; reset temporal history on both switches.
152
+ 6. **The drop lands on an airborne slow hold:** a few high-contrast frames, a short radial zoom, then about 68 frames at
153
+ sim speed 0.2 with a slow push.
154
+ 7. **A one-take reveal with a foreground wipe:** a spline camera past a large foreground object that covers the lens for
155
+ 20-30 frames, then onto the hero as a small figure.
156
+ 8. **Landing in a long lens:** 85-135 mm equivalent at knee height, side-on, the subject sharp, a dust burst, a slow push.
157
+ 9. **Sub-second cuts on motion:** a near-still macro of gear, an ankle-height lateral track at 15 px/f or more (60 fps), an
158
+ overhead tilted wide on the takedown at sim speed 0.25.
159
+ 10. **A title found by a whip pan:** extruded letters at two depths, a 20-frame whip from the running hero onto them, a
160
+ hold, a focus rack between the lines, the hero's blurred body crossing the foreground.
161
+ 11. **A title standing in the scene, lit by it, crossed by the hero** (mesh letters that take the scene's key light and
162
+ cast shadows). The OG Snipers owner wanted theirs floating rather than on the ground.
163
+ 12. **A weapon macro, a sleeve wipe, the next clip:** a macro on a firing sidearm, a whip along the arm, the sleeve
164
+ blacking out the lens for about 16 frames, the next first person already scoped.
165
+ 13. **A punch-in on a clip's last kill** for a few frames around the shot; once or twice in the film.
166
+ 14. **A lens iris of 2-6 frames** to leave a cinematic for gameplay.
167
+ 15. **A tint flash with particles on a special kill:** grade toward one accent from the kill frame, decaying over 8-20
168
+ frames, with the kill's own debris; one colour per kill type; never before the kill frame.
169
+ 16. **An ultra-wide tilted dive:** 14-16 mm at step height, rolled 20-30°, sim speed 0.25 for about 32 frames at the apex.
170
+ 17. **The kill from the victim's side,** from the same simulation take: a camera under and beside the victim, ultra-wide,
171
+ orbiting, sim speed 0.2 as they leave the ground.
172
+ 18. **A pop-out frame:** matte the shot between two white bars and let one bright element render over a bar; once.
173
+ 19. **The weapon run:** three short macro moves along the held weapon, the last pushing into the eyepiece; the next frame
174
+ is first person already scoped on a kill.
175
+ 20. **A word or medal inside the scope:** push into the scope, a white disc with the target as a silhouette and a large
176
+ word behind it, a slow zoom out, a squeeze to a line.
177
+ 21. **Fire into white and recover** over about 20 frames while the player is already moving.
178
+ 22. **A floor orbit ending on a boot over the lens,** then the next clip rising out of the dark.
179
+ 23. **Lightning on the shot:** the scope to white for a frame, arcs around the muzzle that light nearby surfaces, a fall to
180
+ dark.
181
+ 24. **A bullet camera for a collateral:** replay the shot from the bullet with every material switched to matte white
182
+ and ambient occlusion, rushing between victims at a slowed sim speed so bodies react in frame. Once.
183
+ 25. **One held scope, then the shot,** after fast kills: about 52 frames at sim speed 0.2 with a slight sway, the shot on a
184
+ beat. Only if the owner allows held scopes.
185
+ 26. **An iris that becomes the next weapon's scope ring.**
186
+ 27. **The breath, then a backflip handed to first person and back:** a ground-level wide with a slow dolly-in, a chest-up
187
+ shot with something out of focus behind, then a slide, a backflip at sim speed 0.25, an iris into first person, a
188
+ profile close-up in the air while the shot charges, a short strobe on the shot.
189
+
190
+ ## 6. The finishing layer, separated from the game
191
+
192
+ The HUD is drawn in screen space, so it cannot move, grow, tilt or smear unless the edit did it. Fitting each HUD corner
193
+ against its own median template per frame gives the edit's zoom, roll and shift; the HUD's edge energy gives the edit's
194
+ blur. It cannot tell an exposure key in the edit from the game's own muzzle light, and it fails on the most smeared
195
+ frames, so the peak values below are under-read.
196
+
197
+ - **The grade:** contrast is the gap, not exposure. FAINT's first person: mean luma 94, quartiles 29 and 157, black at
198
+ codes 3-5, white reaching 242-247, 14.5 % under code 18, 1.25 % over 245, a teal cast (a* −10.8). Per-clip medians run
199
+ from 48 to 123: it shares the curve and the cast, not the level. The OG Snipers cut before its look test: the same mean,
200
+ quartiles 47 and 127, nothing black and nothing white. No vignette, no visible grain, no chromatic split. **Recipe:**
201
+ levels from the cut's own measured black and white to codes 4 and 246, an S-curve of slope 1.20, a shoulder from 0.80,
202
+ a fraction of the cast in the shadows only (keep places and skins as themselves). OG Snipers then set per-clip exposure
203
+ into a median luma of 86-105; that range is its own, not FAINT's.
204
+ - **The shot accent:** 46 % of shots flash (+25 levels or more, median +49, gone by half in 3-4 frames), 18 % bump, 36 %
205
+ get nothing: not every shot gets the same accent. The zoom punch: a median +3.2 % (p90 +4.8 %), arriving in one frame on
206
+ the shot frame and easing out over about 20 frames at 60 fps; roll a median 0.72°; a damped shake of a few frames. No
207
+ chromatic split. **Recipe:** a class per shot from those shares (flash 40 %, punch 32 %, bump 16 %, clean 12 %), never
208
+ the same class twice running; zoom +3.2 % (+4.8 % on a clip's last kill or a multi-kill); roll 0.7°; a 6 px damped
209
+ shake. The OG Snipers owner then asked that the punch hold until the next payoff ("it goes back to 100% too quick").
210
+ - **Whips:** in a pan whip the world streaks and the weapon and HUD stay sharp: a blur of the world only, not of the whole
211
+ frame. **Recipe:** directional blur of the world along the view's own motion from the capture log, 0.30 of the frame's
212
+ travel, capped at 72 px, only above 30 px of travel a frame and at sim speed 0.6 or more, never from 2 frames before a
213
+ shot to 7 after, with the weapon, hands and HUD matted out. Leave it off until the matte is clean on every skin.
214
+ - **Transitions, by share of FAINT's 68 cuts:** plain cuts on motion 44 %; whips through the cut 16 % (where the picture
215
+ already pans 26 px/f or more at 30 fps; at most 7); white flashes 9 % (1-2 frames; at most 3 where a muzzle frame hands to
216
+ its own outside camera); light-streak zooms 6 %; through the scope 6 % (when the incoming clip opens scoped); foreground
217
+ wipes 4 % (from the take, never pasted); dark beats 4 %; burns to white 3 % (into the ending, once); ink flashes,
218
+ irises and portals 16 % together (of their moment, and easy to date).
219
+ - **No frame blending, no dissolves.**
220
+
221
+ ## 7. The seams of the cut the owner praised for flow
222
+
223
+ The cut the OG Snipers owner later named as the model for flow had no post transitions at all. Its seams cut on the
224
+ game's own leaving motion (the unscope, an aim release, a camera pull-out) into the payoff cinematic, entered first person
225
+ on the frame its swing began, and took foreground wipes from the takes themselves. When a later round stripped the caused
226
+ transitions along with the pasted ones, the owner said "flow/transitions got worse". Measure the seams of every cut the
227
+ owner praises (white and dark frames, view speed out and in, radial push, luma) and keep that table as the model.
228
+
229
+ ## 8. A good player, measured
230
+
231
+ Before scripting a player, record 3-5 minutes of a good human in the real game (the owner, if they play) and measure it.
232
+
233
+ **Getting view angles from a video.** Track features with Lucas-Kanade (`cv2.calcOpticalFlowPyrLK`) on a 960x540 proxy,
234
+ fit a similarity transform per frame pair with RANSAC (`cv2.estimateAffinePartial2D`), with the HUD and the first-person
235
+ weapon masked out, and convert the rotation to degrees through the game's field of view. Treat every absolute angle as
236
+ ±20 % unless you know the player's actual field-of-view setting (ratios, durations, frame counts and curve shapes do not
237
+ depend on it); strafing is not separated from rotation by this fit; frames lost at scope-in and cuts are normal. Find
238
+ shots from the HUD (an ammo counter, a damage number, the scope overlay), not from the audio, which carries every other
239
+ player's gunfire. Measure the build, not its config: OG Snipers' on-screen scope-in took 0.10-0.20 s where the config said
240
+ 0.30-0.40 s.
241
+
242
+ **OG Snipers' owner, measured:** scope-in to full overlay a median 0.13 s; overlay to shot 0.32 s; shot to unscope 0.10 s;
243
+ shot to next shot 1.68 s (quartiles 1.25 and 2.27); hip flicks a median 14.5° in 0.13 s; the view dead still on 1.6-2.1 %
244
+ of hip frames, never more than 4 in a row; a 91 % hit rate (the film needed 100 %).
245
+
246
+ **Bot tells, with thresholds from that play** (fail a take when):
247
+
248
+ 1. A frozen hip camera: over 5 % of frames static, a static run over 6 frames, or any 1 s window under 1° of path.
249
+ 2. Constant-speed turns: any flick whose speed varies less than 20 % (coefficient of variation), or 4 or more frames in a
250
+ row within ±5 % of each other.
251
+ 3. Linear ramps: the flick's speed peak in its first 15 % or last 10 % on more than a fifth of flicks, or progress that is
252
+ a straight line (r² over 0.995). Human flicks peak at about 0.56 of their duration.
253
+ 4. Instant snaps: any single-frame turn over 17°, a peak over 450°/s, or a flick shorter than 60 % of the fitted duration
254
+ (duration ≈ 0.046 × amplitude^0.45 s).
255
+ 5. Zero settle: no overshoot in a clip of 5 or more flicks (40 % of human flicks overshoot, a median 13.5 %), or the scope
256
+ coming up on the flick's last frame every time.
257
+ 6. A dead-still scope at the shot on every shot of a clip (under 0.05° over the last 0.1 s).
258
+ 7. Identical timing: kill gaps varying less than 25 %, two consecutive gaps within 2 frames, or every scoped time within 2
259
+ frames.
260
+ 8. Pure-yaw turns: more than 20 % of flicks with under 0.05° of pitch.
261
+ 9. Always the same scope time across a clip.
262
+ 10. Pixel-perfect centre hits: a human lands up to half a body width off centre; review by eye until the build can
263
+ measure it.
264
+ 11. No strafing, or robotic stop-starts: not measured on OG Snipers; review by eye.
265
+ 12. No turn after the kill: under 2° of turn in the 0.5 s after every kill.
266
+
267
+ **These tells are measured in game time, on the take's own view log. On screen, the owner's rules win:** where they want
268
+ payoffs at equal spacing on the beat, the clock and the payoff solver produce that spacing, and tell 7 does not apply to
269
+ the picture; where they ban a play style (OG Snipers banned held scopes), tells 6 and 9 do not ask for it. The other
270
+ override is deliberate: a human misses, and the film never does.
271
+
272
+ **A human input synthesiser,** sampled per event, never once per clip:
273
+
274
+ | Parameter | Value (OG Snipers) |
275
+ |---|---|
276
+ | Flick duration | 0.046 × amplitude^0.45 s, times a factor of 0.75-1.35; never under 0.07 s |
277
+ | Flick peak speed | 35.4 × amplitude^0.55 °/s, capped at 445°/s |
278
+ | Flick shape | an asymmetric bell, its peak drawn from about 0.56 ± 0.16 of the duration, clipped to 0.30-0.80 |
279
+ | Pitch | pitch = yaw × a ratio of 0.04-0.21, either sign; never exactly 0 |
280
+ | Overshoot | on 40 % of flicks, 3-24 % of the amplitude (median 13.5 %), returned within 0.2 s |
281
+ | Aim error left for the scope | 0.5-2.3° off (median 1.0) |
282
+ | Pause from flick end to scope-in | 0.27-0.58 s (median 0.42); floor 0.07 s |
283
+ | Scoped time before the shot | a mix of short, medium and long, spread at least 0.2 s across a clip (within the owner's play-style rules) |
284
+ | In-scope correction | one move at 20-45°/s peak, ending 0-0.13 s before the shot; leave 0.1-0.4° of motion in the last 0.1 s on a quarter of shots |
285
+ | Release after the shot | unscope at +0.07 to +0.13 s |
286
+ | Weapon swap (two-weapon clips) | input at +0.17 to +0.33 s (median 0.30) |
287
+ | Turn-away | a new flick within 0.5 s; median 14°, up to 58° |
288
+ | Time between kills (when the owner's grid does not fix it) | 0.87-2.5 s, median 1.7 s, coefficient of variation at least 0.35 |
289
+ | Hip idle | never still: 4-8°/s median drift, 0.12-0.19° per-frame jitter at 30 fps, a flick about once a second |
290
+ | Scoped idle | 1.7-3.2°/s median drift, 0.05-0.08° per-frame jitter at 30 fps; still runs up to 0.4 s |
291
+ | Not measured | the reaction delay to a new target, the strafe pattern, jump, slide and flip timing |
292
+
293
+ Measure your own genre's equivalents (a racing line, inputs against frame data, hand paths in VR) the same way.
@@ -0,0 +1,64 @@
1
+ # The song: measuring it into a film map
2
+
3
+ In a montage the owner's song is the clock and the lead. It is played, not edited: one gain from start to end and, at
4
+ most, a fade over the last half second. Measure it before any shot, and write the film map from the measurement.
5
+
6
+ ## 1. Rights and choice
7
+
8
+ - The owner's own track wins. Confirm it is cleared for where the montage will be posted (a commercial song on a
9
+ personal upload is not cleared for a game's marketing). Never assume that music made for a game or offered on a
10
+ platform is cleared for that game's montage.
11
+ - With no track: recommend 2-3 measured candidates with a 15 s excerpt each and a default, from music they own or
12
+ licensed, or from a licensable library with the licence saved beside the file. Generated music runs only on the
13
+ person's own account after they approve the cost; record the approval word for word.
14
+ - A rhythm game whose own chart is not the owner's song: chart the song in the game if it supports custom maps, or mute
15
+ the game and align only shared downbeats.
16
+
17
+ ## 2. Measure it like a music editor
18
+
19
+ Never play audio aloud; take every number from a script (the measuring chapter has the recipes).
20
+
21
+ 1. Decode to 32-bit float WAV; record the decode convention (MP3 adds about 25 ms of priming; a transcode can shift
22
+ everything by about 1.5 frames).
23
+ 2. Duration from the sample count; integrated loudness, true peak and range; momentary loudness every 100 ms.
24
+ 3. **Tempo from attacks in one band** (hi-hat or snare), fitted by least squares; check half and double tempo. Never trust
25
+ a single beat tracker.
26
+ 4. **Check for splices** if the file might be an edit of a longer song: cross-correlate pieces against the full song and
27
+ build a piecewise clock if pieces sit at different offsets. The OG Snipers song ran on one clock with no splice; the OG
28
+ Clash trailer track was three pieces.
29
+ 5. The downbeat phase from section changes; a bar map with each bar's loudness, sub level, air, onsets and chroma.
30
+ 6. Sections: verses and choruses, where any vocal starts and ends, holes and dips, the loudest moment, the key, whether
31
+ the song has an ending, a drop or risers.
32
+ 7. Write `research/grid.json` with beats, bars and sections in seconds and 60 fps frames. Shot specs and the edit read
33
+ positions from it; nothing computes frames from BPM.
34
+ 8. **Click check:** the track with clicks on downbeats and beats; measure its attacks within ±15 ms of each grid beat.
35
+ Every downbeat you cut on must sit within a frame.
36
+ 9. Note the master level: commercial tracks are hot. The OG Snipers film sat at −12.9 LUFS with the song bus at −14.4.
37
+
38
+ OG Snipers: 137.370 BPM on one clock; a beat of 26.207 frames at 60 fps; bar 1 beat 1 at 0.755 s (frame 45); four 16-bar
39
+ blocks of 27.954 s; short-term loudness −10.6 to −8.8 LUFS (flat), with no drop, no riser and no ending; a half-time
40
+ feel with the snare on beat 3.
41
+
42
+ ## 3. The film map
43
+
44
+ - **Map the owner's spoken times onto measured downbeats** and confirm the length in one line. "around 57 seconds" became
45
+ 56.663 s (frame 3400), where the hook returns. The rap ended at 84.44 s and the next downbeat, 84.616 s (frame 5077),
46
+ became the door to the end shot. The first brief said both "around 60 seconds" and "around 1:30".
47
+ - **Slots of 8, 12 or 16 beats,** each with its grid. A slot's last 2 beats are its last payoff's hit and outside shot,
48
+ or the next clip's entry.
49
+ - **Sections drive content:** verse = play, chorus = spectacle. A flat loop has its sections made by the edit and the
50
+ mix. Ambient or orchestral music is cut on phrases, not a grid.
51
+ - **Payoff spacing comes from read time** (the grammar chapter): the smallest power-of-two number of beats that covers the
52
+ payoff's read time, the same beat positions in every clip. Half-time feel: payoffs on the snare, motion peaks on the
53
+ kicks. Drum and bass: treat it as 85-87 BPM for payoffs.
54
+ - **The ending:** one shot of 3-5 s from the door the song gives (the vocal's end), the song fading under it untouched.
55
+
56
+ ## 4. The song in the mix
57
+
58
+ - The music leads. Payoffs sit about 3 dB over the song in 1-5 kHz, not 6 or more. The song is never ducked by accident,
59
+ never edited, never dropped to silence ("u cant have music just go completely blank").
60
+ - Check on the delivered file that the song is untouched: align the song file to the music bus and fit one gain in 1 s
61
+ windows; the gain must hold within 0.5 dB everywhere but the agreed fade.
62
+ - Callouts sit under the song, just after their trigger; test each over the real song at several offsets with two speech
63
+ recognisers before the mix.
64
+ - Before any cut is sent, send the owner a 20 s mix check and ask whether the music is the loudest element.
@@ -0,0 +1,139 @@
1
+ > Carried from `opengame-multiplayer` 1.1.0. Read it when: Build multiplayer for a browser game, or a lobby where bots stand in for players. Covers one authority deciding what happened, a relay that only forwards and polices, capped messages, host migration and rejoin, smooth remote players, the lobby as the loading screen, a match that never pauses, bot lobbies, fair spawns, voice, testing several browsers and naming what was never proven. Use when the person asks for co-op, online play, private lobbies, voice chat or bots that feel like players.
2
+
3
+
4
+ # Multiplayer
5
+
6
+ This skill is for a browser game that more than one person plays at once, and for the
7
+ single-player game that has to feel like one: a lobby filled with bots presented as
8
+ players. It comes from two games and an honest gap. LC built four-player private lobbies,
9
+ with rooms on a Cloudflare Worker and Durable Objects, and measured them on headless pages;
10
+ they were never deployed, nobody played them by hand, and no voice was ever spoken or
11
+ heard through them. OS filled lobbies of up to 32 with bots and built no networking at
12
+ all; the owner judged it by whether it felt online. So the rules say what was measured,
13
+ and where nobody has been yet, and every report you write about multiplayer does the same.
14
+
15
+ Rules carry five tags; `opengame-quality-bar` explains them and the game codes. LC's
16
+ protocol, its liveness and rejoin numbers, smoothing code and the test rig are in
17
+ [references/netcode.md](references/netcode.md). OS's simulated lobby (arrivals, churn,
18
+ names, caps) is in [references/bot-lobbies.md](references/bot-lobbies.md).
19
+
20
+ **OpenGame does not run your rooms.** The OpenGame SDK has no rooms, lobbies or relay; its
21
+ multiplayer module was removed, so do not build on room calls left in older integrations.
22
+ The game brings its own server. `opengame-deploy` covers how the game reaches it: the
23
+ game's exact origin on the server's allow-list, a token in a header or the first socket
24
+ message and never a cookie, and a first frame that still draws when the server is out of
25
+ reach. `opengame-sdk` gives each player a name and an OG for the lobby, and the invite link
26
+ (`og.invite`) that opens the game on OpenGame. The full guide, with server settings, is at
27
+ https://docs.opengame.io/guides/multiplayer. How bots move, aim
28
+ and get unstuck is in `opengame-enemies`; the lobby's shown gate and byte budgets are in
29
+ `opengame-loading`; the press that starts play and pointer lock are in `opengame-front-end`.
30
+
31
+ ## Rules
32
+
33
+ ### An invite opens the game on OpenGame and seats the friend
34
+ - **Invariant:** Every invite or lobby link the game shows is the game's page on opengame.io with `?join=CODE`, and a player who opens one lands in that lobby with no menu in between. When the lobby is gone or full, the game says so and offers a new lobby or solo play.
35
+ - **Technique:** Build the link with `og.invite.link(code)` and show the code beside it, since `link` returns null outside OpenGame and for a code the page would change. After `await og.init()`, a set `og.invite.code` means the player opened an invite: join that lobby at once. Use 4 to 12 capitals and digits without look-alikes for codes people read out, or 22 characters of `A-Z a-z 0-9 _ -` for an id they only copy. Never link to the game's `<label>.opengame.gg` address: opened in a tab it shows a page pointing to OpenGame and drops the code. A code is not a password, so a private lobby has the host accept players or locks once full.
36
+ - **Reference configuration:** The owner asked on 16 Nov 2026 that "the multiplayer lobby link is opengame so people who join get put right into lobby on opengame". OG HERO's six-letter rooms link to its game page with `?join=CODE`, and the page starts the game as it opens on a computer; on a touch screen the player presses Play first and the code still arrives.
37
+ - **Applies when:** Any game with lobbies, rooms or private matches.
38
+ - **Verify:** Two separate browsers through opengame.io: one hosts and copies the invite, the other opens it and is in the host's lobby within seconds, with no menu pressed.
39
+
40
+ ### PeerJS runs on servers the creator chose
41
+ - **Invariant:** A published PeerJS game signals through a PeerServer the creator runs over `wss:`, reaches players behind strict routers through TURN, keeps TURN credentials out of the bundle, and uses peer ids no other app can take.
42
+ - **Technique:** Run `PeerServer` from the `peer` package behind TLS with `allow_discovery: false`, `corsOptions` naming the game's exact origin, and an `Origin` check in front of its socket (`opengame-deploy`). Pass `host`, `port: 443`, `secure: true`, `key` and `config.iceServers` to `new Peer`. Use STUN plus a TURN provider whose short-lived credentials the creator's server issues. Prefix peer ids with the game's name; PeerJS ids are letters and digits with single `-`, `_` or spaces between, so use short codes. On host loss, the next host retries for the lobby's id until the signalling server frees it.
43
+ - **Reference configuration:** The public `0.peerjs.com` is shared by every PeerJS app with no uptime promise. OG HERO's host migration lost players until the new host retried its claim once the old host was gone and re-sent state every two seconds (four browsers, the host killed mid-song). `opengame check` warns on each STUN or TURN server a game names and notes the PeerJS defaults.
44
+ - **Applies when:** Games that connect players with PeerJS or raw WebRTC.
45
+ - **Verify:** Kill the host's browser mid-match: another player hosts and an invite shared before still joins. Join from a phone network: the connection goes through TURN.
46
+
47
+ ### Players send what they did; one authority decides what happened
48
+ - **Invariant:** Each player's own movement and shots answer on their own page at once and never wait on the network. Every outcome that matters (damage, kills, pickups, points, enemy state, doors) is decided by one authority, the host page or a server, and a client's report of a hit is a claim that authority checks, never a result it accepts.
49
+ - **Technique:** Generate the map from a shared seed. Each page simulates its own body and sends its pose and inputs; a shot goes out with its origin, direction and time, and the authority tests it against its own copy of the world (line, range, rate of fire, what stands in the way) before applying damage. The authority owns every shared object and the rules. Scores the game reports to OpenGame stay the game's claim (`opengame-sdk`).
50
+ - **Reference configuration:** LC: the map comes from the seed; each player's body and shots belong to their own page; the host page owns the zombies, the tide, the stations, doors, points and weather; a hit claimed 40 m off the round's line was refused on four pages. OS ran its whole lobby on one page and gave every bot the human input path, recoil, scope and bolt, so a bot's shot went through the same rules as the player's.
51
+ - **Applies when:** Co-op and competitive games with shared state.
52
+ - **Verify:** A test client sends a hit 40 m off its line, a kill it did not make and a pickup out of its reach; the authority refuses each and logs why.
53
+
54
+ ### The relay forwards and polices; it never simulates
55
+ - **Invariant:** The relay passes messages between pages, keeps the host's latest checkpoint and enforces limits, and runs no game logic. Every message kind has a declared size and rate shared by page and relay, a client that breaks them is closed, and nothing over a limit is dropped without a count and a log line.
56
+ - **Technique:** One protocol module imported by both page and relay, naming each message kind with its byte size, rate and direction. The relay counts fouls and closes a line after a few, refuses oversized frames and floods, and logs every drop. A capability number in the welcome message lets two builds share a room or turns the older one away with a reason. Private rooms take short join codes from an alphabet without look-alike characters, with a lockout after repeated wrong codes. Keep each room's state in one object of its own, so a room outlives a page reload.
57
+ - **Reference configuration:** LC's relay is a Cloudflare Worker with two SQLite Durable Objects that "never simulates anything". A body update is exactly 28 B at 30 a second; events are 6-32 B; the host's checkpoint is up to 4096 B about once a second, kept by the room; the host's world update is up to 1200 B at 80 a second (at an earlier cap of 32 a second, the room "dropped what was over without a word"); a client's request to the host is up to 29 B; chat is up to 200 characters. Eight fouls close the line, as does a frame over 4200 B or more than 300 frames a second. Join codes are 6 characters, and 10 wrong codes lock an address for a minute. With three players a guest received about 5,100 B/s and the host sent about 4,925; the relay billed about 13,000-19,000 requests a room-hour.
58
+ - **Applies when:** Any game with a relay or room server.
59
+ - **Verify:** Fuzz the relay from a test client with oversized, malformed and over-rate frames: each is refused or closes the line, and each drop appears in the log with its kind.
60
+
61
+ ### The host can leave at any moment
62
+ - **Invariant:** When the host's tab closes or goes silent, a standing player becomes host almost at once and the world keeps moving; a stall on one page's main thread drops nobody; and a player who reloads rejoins their own seat with what they had.
63
+ - **Technique:** Run the socket in a Web Worker, so a long frame on the main thread cannot silence it. Detect silence on both the room and the page within seconds, retry with a short backoff, and hold a seat for minutes, longer than a page's own boot on a busy machine. Migrate on the close code for a page going away or on an away message sent from `pagehide`, to the lowest seat still standing, from the checkpoint the room keeps. Keep a rejoin token in the tab and in the browser with a Web Lock per seat. A reloading player is not drawn and cannot be hit until their page says it is playing, and respawns clear of enemies with a moment of grace.
64
+ - **Reference configuration:** The owner asked that "host migrations should come out of box too if host leaves" and that "no one disconnects randomly as a client". LC: the room declares silence at 2 s and the page at 3 s; retries at 0, 0.2, 0.4, 0.8 and then every 1 s; a seat is held 2 minutes, because 30 s was shorter than one page's boot on a busy machine; migration fires on close code 1001. The host was named 49-91 ms after a close, and zombies stood still at most 26 ms where they had stood 1.8-2.0 s; the first version handed the world to a downed player ("the night passed to a man who was out"), hence the standing rule. The rejoin token lasts 2.5 minutes, and a reload rejoins before the title with position, weapons and rounds kept. Respawn is at the player's own spot when no zombie is within 5 m, otherwise the nearest spot clear for 8 m, facing the nearest zombie, with 1 s of grace.
65
+ - **Applies when:** Games where one player's page hosts, and any room that must survive a player leaving.
66
+ - **Verify:** Close the host's tab mid-fight: a new host within 0.15 s and enemies still for less than that. Freeze a guest's main thread for 5 s: still seated. Reload a guest: back in the same seat with the same kit.
67
+
68
+ ### Every player sees the same world, and other players move smoothly
69
+ - **Invariant:** What one player sees of shared things (which enemy model and variant, where another player stands and what they hold) matches what every other player sees, and other players move without jumps, pops or late arrivals.
70
+ - **Technique:** The authority picks cosmetic variants only from assets every seat has loaded, and sends its choice. Load every other player's model and gear in the lobby. Draw remote players a fixed interval behind, estimate each sender's clock offset from the smallest arrival delay over a short window, and extrapolate only a little past the last update. Measure drawn error on real pages, because a simulation of the smoothing can show gains that real pages do not. Spectating runs the same camera code as playing, from the watched player's position.
71
+ - **Reference configuration:** The owner asked for "perfectly synced and how players look on each others screens and how they move hold guns get hit go down (and can be revived)". LC drew remote players 100 ms behind, took the clock offset as the minimum of arrival minus send time over 2 s, and extrapolated at most 150 ms; drawn error had a median of 0.5 cm and a p95 of 18-22 cm, and 87 of 102 tail samples came from the sender's own long frames, "not the line". A simulation said one smoothing change took the p95 from 84 to 27 cm; real pages showed 71.7 against 71.8. Before cosmetics came from shared assets, 7 of 18 zombies were a different person on a guest's page; after, 18 of 18 matched. Remote players were first seen 5.15-6.05 s into a run until their models loaded in the lobby (then 0.16-0.19 s). Spectating sat 0-2 mm from the watched player's own lens.
72
+ - **Applies when:** Any game where players see each other or shared enemies.
73
+ - **Verify:** Screenshot one moment on every seat and compare variants and positions; log drawn error per remote player on real pages through a fight.
74
+
75
+ ### Every seat is loaded before the countdown ends
76
+ - **Invariant:** The countdown ends only when every seated player's page has the map shown and every other player's model, gear and finishes loaded, so nobody appears late, empty-handed or as a stand-in, and the first minute of play downloads nothing new on any seat.
77
+ - **Technique:** Each page reports to the room when its map passes the "shown" gate and when every other seat's model and kit have loaded; the countdown waits for the slowest seat, with a visible wait and a cap after which a late seat joins as a loading player who is not drawn and cannot be hit. A player who joins mid-match loads everyone in the lobby screen before dropping in. The shown gate itself, the order of the download line and its budgets are in `opengame-loading`.
78
+ - **Reference configuration:** LC's remote players were first seen 5.15-6.05 s into a run until their models loaded in the lobby, then 0.16-0.19 s; "the megabytes move and do not grow". OS's deploy waits for every seated rifle and optic, and moved 19.6 MB of killstreak art, its largest first-match download, into the countdown; its owner had reported "it says 10 seconds, then 5, 4 and then nothing happens and the game just keeps loading?...".
79
+ - **Applies when:** Match-based games with lobbies, real or simulated.
80
+ - **Verify:** On every seat, a request log bucketed by phase shows nothing fetched in the first minute of play; a throttled seat holds the countdown, and no other player is ever drawn without their model.
81
+
82
+ ### Opening a menu never pauses the match
83
+ - **Invariant:** The game's own menu, settings and idling never stop the shared match: other players, bots, timers, kill cams and spawns go on, the world's sound carries on ducked behind the menu, and a player in the menu stays in the world and can be hit. When the OpenGame page pauses the game, a networked page stops its own input, sound and drawing while the shared match runs on at the authority; a single-player game, bots included, pauses as the SDK says.
84
+ - **Technique:** Keep the simulation clock independent of the interface; the game's menu only captures input. In `og.onPause(true)` after play has started, a networked client stops sending input and mutes, and a host page hands the world to another standing seat rather than stopping it for everyone (the pause a page holds under its poster before Play is not a hand-off: that page has never held a world); a game with no network stops its simulation, timers and audio (`opengame-sdk`). Revive and bleed-out timers run on wall time. The press that starts play, pointer lock and spawning without a "click to take control" step are in `opengame-front-end`; inside OpenGame, fullscreen goes through `og.fullscreen.request()`.
85
+ - **Reference configuration:** OS's owner asked six times, among them "everyone needs to keep playing -> pause is just u opening menu not pausing the game!" and "u should hear the gameplay in background (muted a bit) but not just the voices"; in OS a visible paused player can be killed. LC's revive rings and bleed-outs run on the wall clock in co-op. Neither game measured a host page being hidden or paused by the page, so treat that path as untested.
86
+ - **Applies when:** Online games, co-op and bot-filled games presented as online.
87
+ - **Verify:** Film 10 s of the game's own menu open mid-fight: others move, timers run, the world is audible. Hide the host's tab: another seat takes the world, and the enemies keep moving.
88
+
89
+ ### Bots standing in for players fill a lobby that reads as online
90
+ - **Invariant:** A lobby of bots presented as players arrives, leaves and behaves like one: it fills gradually and not always to the cap, some leave during and after the match, names look like real players' handles, nobody fakes speech, bots get no powers a person would not have, and the game says somewhere players can find it (the store page, a settings line) that bots fill empty seats.
91
+ - **Technique:** Draw arrival times from a long-tailed distribution and fill to the cap only some of the time; let a share leave during the match with a killfeed line, and most stay for the next. Generate handles in the style of real platforms, mostly lowercase, a minority with digits or underscores. Leave a bot's mic slot empty. Cap bot killstreaks and powers on the match clock. Once kill cams or replays exist the player watches bots closely, so give them the human input path and timing (`opengame-enemies`).
92
+ - **Reference configuration:** OS's owner: "it should feel like ur playing online multipalyer wiht real people". Arrivals are lognormal, filling to exactly 32 only 70% of the time ("always exactly 32 looks scripted"); 5-20% churn shows in the killfeed after "we should have bots leave the game throughout (not many of them maybe 10% of them)"; about 80% stay after the match; an empty mic slot ("never fake talking"). Handles: "realistic handles from playstation/steam/xbox ... mix of all lowercase" with numbers "maybe 10-15%" (owner); the generator lands 12.7% with digits or underscores, held after the third retune. Bots get only a counter-UAV, "at most once every 5 minutes for at most 30 seconds" (owner), and one shield per 5 minutes; a jam spaced on match time but timed on frame time lasted 90 s, not 30. Kill cams made bots visible, so no out-of-sight teleport and no one-frame head lock ("tells the player it was a bot").
93
+ - **Applies when:** Games that fill seats with bots presented as players.
94
+ - **Verify:** Log arrivals, departures, names and mic states over ten simulated lobbies: fill and churn land in their ranges, and a frame-by-frame look at bot kill cams shows no snap.
95
+
96
+ ### Seats, spawns and quality tiers are fair at every size
97
+ - **Invariant:** The seat count follows the map's area and the frame budget, no player spawns in a rival's sight, and no quality tier lets one player see through cover that hides them from another.
98
+ - **Technique:** Cap a lobby by area per player. Gate spawns on line of sight and never relax the gate: glass, foliage, rails and lattice count as see-through, and fog does not hide. Scale spawn distances to the map rather than using one ideal distance. Measure frame cost per live body, so a drop in the first minute is read as the lobby filling. Compare quality tiers in a gameplay-only A/B of cover, sightlines and doors.
99
+ - **Reference configuration:** OS capped lobbies at 32 on big maps and 14 on the smallest; "Glass, foliage, rails and lattice count as see-through, because snipers shoot through what they can see through"; a fixed 40.6 m ideal spawn distance saturated on small maps. Frame time scaled with live bodies: 3.0 ms with one, 14.4 ms at 24, up to 39 ms at 32. A low tier that thinned tree crowns let "a weak-machine player see through cover that hides players on normal machines".
100
+ - **Applies when:** Competitive games, real or with bots, and any game with quality tiers.
101
+ - **Verify:** A spawn audit per map with every spawn's sight lines to every other; a frame-time table by live bodies; a tier A/B of the same sightlines.
102
+
103
+ ### Voice opens on the player's press and lives in the world
104
+ - **Invariant:** The microphone opens only on the player's own press and shows when it is live; proximity voice falls off over a long range and through walls; the game makes room under a voice; players know a peer-to-peer call shares their IP address with the others; and nothing about voice is called working until people have talked through it.
105
+ - **Technique:** A WebRTC mesh carries a few players. Past that, or when players must not see each other's addresses, relay voice through a server (TURN or a forwarding server). Ask for the mic from a click, never at load. Route each remote voice through the game's spatial audio: gain by distance with a smooth roll-off to silence, a muffle through walls, and a light duck on the game's mix while anyone speaks. Let the host choose proximity or everyone, and keep a mute per player. Prove the graph with a tone fed in at known distances, and write in the report that no voice was heard.
106
+ - **Reference configuration:** LC (designed, never heard): a WebRTC mesh with Opus and STUN only and no TURN server, so a player behind a strict NAT would have had no path, and nobody tested one. The mic opened only on the player's press; proximity gain was 0.7 at 40 m, then a raised cosine to 0 at 90 m; through a wall x0.6 and a 900 Hz low-pass; the game ducked 4 dB under a voice; text chat took up to 200 characters. The owner had asked that players "can choose proximity or omni chat (host) for voice chat" and that the "master volme of game may need to come down a bit and make sure proximity chat has a large range and is well implemented". OS left every bot's mic slot empty.
107
+ - **Applies when:** Any game with voice or text chat.
108
+ - **Verify:** A tone injected as a remote voice and measured at the output at 10, 40 and 90 m and behind a wall; the mic stays closed until a real press; the report names what nobody has heard.
109
+
110
+ ### Test several real browsers in one job, through a bad network
111
+ - **Invariant:** Multiplayer claims come from real game pages, several in one test, through a network that delays, jitters and loses packets, with each measuring tool proven and at least one player driven by real input events.
112
+ - **Technique:** Run all the pages of one test as a single job in your machine's job queue, so the browsers share one slot rather than each holding one and waiting for another. Put a relay that delays, jitters and holds packets between the pages and the room server, because the browser's own network throttling does not hold WebSocket frames. Count frames at the draw call, not at animation callbacks. Give test players a cheat that only scripts can turn on, leaving one vulnerable so damage is still tested. Never read the canvas back inside a timed window. Press every lobby control with a real input event at least once. Test the room server alone, without a browser, at the protocol's edges.
113
+ - **Reference configuration:** LC: four pages reported "56.9-57.3 fps" from animation callbacks while 17.4-18.5 frames a second were drawn; "no button of the lobby's panel ever answered a real press" until the director pressed one; a canvas read-back froze a timed revive for 3 s; 114 room tests ran with no browser; a load test put two rooms of two on one machine with one player reloading over a 20 Mbit line. Its test pipe ran at 200 ms and 2% loss.
114
+ - **Applies when:** Any multiplayer game tested by agents.
115
+ - **Verify:** A recorded four-page run through the pipe at 200 ms and 2% loss, one player on real input, frame counts from the draw call and every lobby control pressed.
116
+
117
+ ### Say what nobody has done
118
+ - **Invariant:** Every multiplayer report names, at the top, each human part nobody has done yet: a deployed relay, people playing by hand on separate machines and networks, a voice spoken and heard, a connection through a strict NAT, a real controller. Headless pages on one machine prove the protocol, not the experience.
119
+ - **Technique:** Open each handover with the list of what was never done, and remove an item only when somebody has done it, saying who, when and how. Ask the owner for a session with real people on separate networks before calling it ready. Keep numbers from simulations apart from numbers from pages.
120
+ - **Reference configuration:** Every LC net handover opened "NOTHING IS DEPLOYED. NOBODY HAS SPOKEN OR LISTENED. NOBODY PLAYED BY HAND." Its lobbies were built and measured on headless pages, never deployed and never played by people. OS built no networking, and its owner judged the bot lobby only by whether it felt online.
121
+ - **Applies when:** Every multiplayer report.
122
+ - **Verify:** The report's first lines list what is untested, and each item taken off the list names who did it and when.
123
+
124
+ ### The game connects from its own origin and plays without its server
125
+ - **Invariant:** The game reaches its server only from its exact OpenGame origin with a token of its own, draws its first frame and plays alone when the server is out of reach, and tells the player in plain words what failed. A name a client sends is a claim that labels a player and never decides a ban, a reward or a seat.
126
+ - **Technique:** Put the game's origin on the server's allow-list, never a pattern on the games domain, and send a token in a header or the first socket message, not a cookie (`opengame-deploy`). Name players from `og.player.handle`, falling back to a guest name when it is null; key a save on `og.player.id` only when it is a string, and key a seat on the room's own rejoin token (`opengame-sdk`). Start solo play without waiting for the lobby; a lobby that cannot connect says so and offers solo.
127
+ - **Reference configuration:** OpenGame's review sandbox does not reach the internet, and automatic review plays every game (`opengame-deploy`). OpenGame serves each game from its own origin and an experiment's versions from `v-` origins, each of which the server must allow; guests and players who do not share their OG have no handle, and guests have no id. LC's room server and pages only ever ran on one machine, so neither path was tried against a real origin.
128
+ - **Applies when:** Every multiplayer game published on OpenGame.
129
+ - **Verify:** Block the server's host and boot the game: the first frame draws, solo starts and the lobby shows a plain error. Connect from the published origin without a token: refused. Join as a guest: a guest name shows and the seat holds through a reload.
130
+
131
+ ## Before you call multiplayer done
132
+
133
+ Run each and say which you could not run; an unrun check is unverified.
134
+
135
+ - A fuzzed relay, a lying client refused, and the host's tab closed mid-fight.
136
+ - Four real pages through a bad network, one on real input, frames counted at the draw.
137
+ - A request log from boot through the lobby with nothing fetched in the first minute.
138
+ - The game's own menu filmed mid-fight, and the host's tab hidden mid-fight.
139
+ - The list of what nobody has done, at the top of the report.
@@ -0,0 +1,71 @@
1
+ # A lobby of bots that reads as online (OS)
2
+
3
+ OS had no networking. Every seat but the player's was a bot, and the owner's test was
4
+ whether it felt like "playing online multipalyer wiht real people". Most of that feeling
5
+ came from things around the match rather than from aim: who arrives, who leaves, what
6
+ they are called, and whether the match stops when you open a menu.
7
+
8
+ ## Arrivals
9
+
10
+ Real lobbies fill in bursts with a long tail, and are not always full.
11
+
12
+ ```js
13
+ // A standard normal from two uniforms (Box-Muller), then a lognormal delay in seconds.
14
+ const normal = () => Math.sqrt(-2 * Math.log(1 - Math.random())) * Math.cos(2 * Math.PI * Math.random());
15
+ const arrivalDelay = (medianSeconds, spread) => medianSeconds * Math.exp(spread * normal());
16
+
17
+ function lobbyFill(cap, fullShare = 0.7) {
18
+ // OS filled to exactly the cap only 70% of the time: "always exactly 32 looks scripted"
19
+ return Math.random() < fullShare ? cap : cap - 1 - Math.floor(Math.random() * Math.max(1, Math.round(cap * 0.1)));
20
+ }
21
+ ```
22
+
23
+ Seat the bots at their delays while the map streams in the lobby, so the list grows the
24
+ way a real one does.
25
+
26
+ ## Leaving and staying
27
+
28
+ - 5 to 20% of bots leave during a match, each with a killfeed line, after the owner's "we
29
+ should have bots leave the game throughout (not many of them maybe 10% of them)".
30
+ - About 80% stay into the next match; the rest are replaced by new arrivals.
31
+ - A paused or idle human stays in the world and can be killed, as online.
32
+
33
+ ## Names
34
+
35
+ The owner asked for "realistic handles from playstation/steam/xbox ... mix of all
36
+ lowercase", "not garbage", with numbers "maybe 10-15%". OS's generator mixes word
37
+ fragments, lowercase by default, with digits or underscores in 12.7% of names, and it took
38
+ three retunes to hold. Check a sample of a thousand: no real person's name, no slur, no
39
+ trademark, no repeats within a lobby. Never reuse the names of real players you have seen.
40
+
41
+ ## Talking
42
+
43
+ Leave every bot's mic slot empty. OS's rule was "never fake talking": a speaking icon with
44
+ no voice, or a canned voice line, gives the bots away faster than their aim does.
45
+
46
+ ## Powers
47
+
48
+ Bots get only what a person would plausibly use, on the match clock. OS's bots got only a
49
+ counter-UAV, "at most once every 5 minutes for at most 30 seconds", and one shield per
50
+ 5 minutes. A jam that was spaced on match time but timed on frame time lasted 90 s, not
51
+ 30, so put every rule on one clock.
52
+
53
+ ## What kill cams change
54
+
55
+ Once the player can watch a bot's view, bots need human timing everywhere: no teleport
56
+ out of sight, no one-frame head lock ("tells the player it was a bot"), reactions with a
57
+ floor, flicks that undershoot and correct. OS's early bots were aimbots: 26 of the
58
+ player's 37 deaths were quickscopes. Aim, movement and getting unstuck are in
59
+ `opengame-enemies`.
60
+
61
+ ## Saying so
62
+
63
+ Players have a right to know who they are playing against. Put a plain line somewhere they
64
+ can find it, such as the store page or a settings screen, that bots fill empty seats. It
65
+ costs the feeling nothing in a match.
66
+
67
+ ## Cost
68
+
69
+ Frame time follows live bodies: in OS, 3.0 ms with one, 14.4 ms at 24 and up to 39 ms at
70
+ 32. A frame-rate drop in a match's first minute is the lobby filling, so size seats per
71
+ map from both its area and that table (OS: 32 on the big maps, 14 on the smallest).