{
  "schema": 1,
  "generated": 1788752047,
  "template": "pipe-dream",
  "title": "Using 1991's 'Pipe Dream' to test LLM one-shot coding tasks",
  "lede": "Each coding agent is launched on a fresh VM with a small number of initial tools, but with full root and internet access. All operate independently with no knowledge of other agents. The only outside influence is a push to denote time and budget remaining",
  "brief": "design a fully-functional, ready-to-play, modern reimagining of the 1991 game \"Pipe Dream\", including a webui that will allow a user to play it immediately. Make the game an award-winning contender, complete with impressive graphics, sound, level progression, high-scores, and appropriate settings.",
  "iframe_sandbox": "allow-scripts allow-pointer-lock allow-modals",
  "ratings_api": {
    "summary": "api/ratings",
    "vote": "api/rate"
  },
  "runs": [
    {
      "id": "1fc170",
      "created": 1788748398,
      "entries": [
        {
          "id": "1fc170-9011",
          "vmid": 9011,
          "name": "duel-1fc170-codex",
          "model": "gpt-6-astra",
          "effort": "",
          "note_app": "Lumen \u2014 a luminous pipe-routing arcade game with campaign, daily challenge, endless and zen modes",
          "commits": 5,
          "ins": 1117,
          "dels": 165,
          "tests": 8,
          "took_s": 2289,
          "rank": null,
          "assessment": null,
          "tokens": {
            "in": 136440,
            "cw": 0,
            "out": 66043,
            "cr": 4454272,
            "behind": false,
            "hi": 0.0
          },
          "cost": {
            "usd": 9.12,
            "exact": true
          },
          "play": "e/1fc170-9011/play/",
          "shots": [
            "e/1fc170-9011/shots/01-live-network.png",
            "e/1fc170-9011/shots/02-sector-complete.png",
            "e/1fc170-9011/shots/03-mobile-workshop.png",
            "e/1fc170-9011/shots/04-core-network.png"
          ],
          "source": "e/1fc170-9011/source.zip"
        },
        {
          "id": "1fc170-9022",
          "vmid": 9022,
          "name": "duel-1fc170-antigravity",
          "model": "gemini-3.8-flash",
          "effort": "",
          "note_app": "PIPE FLUX: Cyber Conduit - Modern award-winning reimagining of 1991 Pipe Dream",
          "commits": 1,
          "ins": 7615,
          "dels": 0,
          "tests": 3,
          "took_s": 684,
          "rank": null,
          "assessment": null,
          "tokens": {
            "in": 534399,
            "cw": 0,
            "out": 105882,
            "cr": 12540475,
            "behind": false,
            "hi": 0.0
          },
          "cost": {
            "usd": 1.74,
            "exact": true
          },
          "play": "e/1fc170-9022/play/",
          "shots": [
            "e/1fc170-9022/shots/01-title.png",
            "e/1fc170-9022/shots/02-mid-game.png",
            "e/1fc170-9022/shots/03-victory.png",
            "e/1fc170-9022/shots/04-mobile.png"
          ],
          "source": "e/1fc170-9022/source.zip"
        },
        {
          "id": "1fc170-9027",
          "vmid": 9027,
          "name": "duel-1fc170-codex2",
          "model": "gpt-5.6-sol",
          "effort": "",
          "note_app": "Fluxworks \u2014 a complete atmospheric pipe-routing puzzle game",
          "commits": 3,
          "ins": 1248,
          "dels": 19,
          "tests": 1,
          "took_s": 748,
          "rank": null,
          "assessment": null,
          "tokens": {
            "in": 72279,
            "cw": 0,
            "out": 25710,
            "cr": 2225152,
            "behind": false,
            "hi": 0.0
          },
          "cost": {
            "usd": 2.25,
            "exact": true
          },
          "play": "e/1fc170-9027/play/",
          "shots": [
            "e/1fc170-9027/shots/01-title.png",
            "e/1fc170-9027/shots/02-gameplay.png",
            "e/1fc170-9027/shots/03-mobile.png"
          ],
          "source": "e/1fc170-9027/source.zip"
        }
      ]
    },
    {
      "id": "ee6f71",
      "created": 1788202730,
      "entries": [
        {
          "id": "ee6f71-9033",
          "vmid": 9033,
          "name": "duel-ee6f71-grok",
          "model": "grok-4.6-build",
          "effort": "low",
          "note_app": "FLOOZ \u2014 modern Pipe Dream reimagining",
          "commits": 5,
          "ins": 1867,
          "dels": 39,
          "tests": 0,
          "took_s": 2786,
          "rank": null,
          "assessment": null,
          "tokens": {
            "in": 443009,
            "cw": 0,
            "out": 21380,
            "cr": 984960,
            "behind": false,
            "hi": 0.0
          },
          "cost": {
            "usd": 1.51,
            "exact": true
          },
          "play": "e/ee6f71-9033/play/",
          "shots": [
            "e/ee6f71-9033/shots/01-title.png",
            "e/ee6f71-9033/shots/02-play.png",
            "e/ee6f71-9033/shots/03-mid-game.png",
            "e/ee6f71-9033/shots/04-mobile.png"
          ],
          "source": "e/ee6f71-9033/source.zip"
        },
        {
          "id": "ee6f71-9034",
          "vmid": 9034,
          "name": "duel-ee6f71-grok2",
          "model": "grok-4.6-build",
          "effort": "medium",
          "note_app": "CONDUIT: modern Pipe Dream reimagining",
          "commits": 7,
          "ins": 3299,
          "dels": 767,
          "tests": 0,
          "took_s": 2794,
          "rank": null,
          "assessment": null,
          "tokens": {
            "in": 324907,
            "cw": 0,
            "out": 68979,
            "cr": 5749248,
            "behind": false,
            "hi": 0.0
          },
          "cost": {
            "usd": 3.94,
            "exact": true
          },
          "play": "e/ee6f71-9034/play/",
          "shots": [
            "e/ee6f71-9034/shots/01-title.png",
            "e/ee6f71-9034/shots/02-play.png",
            "e/ee6f71-9034/shots/03-flow.png",
            "e/ee6f71-9034/shots/04-mobile.png"
          ],
          "source": "e/ee6f71-9034/source.zip"
        },
        {
          "id": "ee6f71-9035",
          "vmid": 9035,
          "name": "duel-ee6f71-grok3",
          "model": "grok-4.6-build",
          "effort": "high",
          "note_app": "CONDUIT \u2014 Pipe Dream Rewired. Neon pipe-laying puzzle.",
          "commits": 3,
          "ins": 8639,
          "dels": 42,
          "tests": 2,
          "took_s": 2343,
          "rank": null,
          "assessment": null,
          "tokens": {
            "in": 462013,
            "cw": 0,
            "out": 89286,
            "cr": 6093312,
            "behind": false,
            "hi": 0.0
          },
          "cost": {
            "usd": 4.51,
            "exact": true
          },
          "play": "e/ee6f71-9035/play/",
          "shots": [
            "e/ee6f71-9035/shots/01-title.png",
            "e/ee6f71-9035/shots/02-mid-game.png",
            "e/ee6f71-9035/shots/03-howto.png",
            "e/ee6f71-9035/shots/03b-settings.png",
            "e/ee6f71-9035/shots/04-sector-clear.png",
            "e/ee6f71-9035/shots/05-mobile-title.png",
            "e/ee6f71-9035/shots/06-level-intro.png",
            "e/ee6f71-9035/shots/07-walls.png"
          ],
          "source": "e/ee6f71-9035/source.zip"
        },
        {
          "id": "ee6f71-9036",
          "vmid": 9036,
          "name": "duel-ee6f71-grok4",
          "model": "grok-4.5-build",
          "effort": "low",
          "note_app": "Pipe Flux \u2014 modern Pipe Dream (1991) remake with glow pipes, flux flow, levels, scores",
          "commits": 10,
          "ins": 4421,
          "dels": 74,
          "tests": 1,
          "took_s": 3595,
          "rank": null,
          "assessment": null,
          "tokens": {
            "in": 285697,
            "cw": 0,
            "out": 43414,
            "cr": 2757504,
            "behind": false,
            "hi": 0.0
          },
          "cost": {
            "usd": 1.66,
            "exact": true
          },
          "play": "e/ee6f71-9036/play/",
          "shots": [
            "e/ee6f71-9036/shots/01-title.png",
            "e/ee6f71-9036/shots/02-gameplay.png",
            "e/ee6f71-9036/shots/03-mid-build.png",
            "e/ee6f71-9036/shots/04-flow-or-result.png",
            "e/ee6f71-9036/shots/05-settings.png"
          ],
          "source": "e/ee6f71-9036/source.zip"
        },
        {
          "id": "ee6f71-9037",
          "vmid": 9037,
          "name": "duel-ee6f71-grok5",
          "model": "grok-4.5-build",
          "effort": "medium",
          "note_app": "Flowline \u2014 neo Pipe Dream",
          "commits": 9,
          "ins": 4952,
          "dels": 188,
          "tests": 0,
          "took_s": 3589,
          "rank": null,
          "assessment": null,
          "tokens": {
            "in": 603270,
            "cw": 0,
            "out": 53410,
            "cr": 3085952,
            "behind": false,
            "hi": 0.0
          },
          "cost": {
            "usd": 2.45,
            "exact": true
          },
          "play": "e/ee6f71-9037/play/",
          "shots": [
            "e/ee6f71-9037/shots/01-title.png",
            "e/ee6f71-9037/shots/02-mid-game.png",
            "e/ee6f71-9037/shots/03-levels.png",
            "e/ee6f71-9037/shots/04-settings.png",
            "e/ee6f71-9037/shots/05-win.png"
          ],
          "source": "e/ee6f71-9037/source.zip"
        },
        {
          "id": "ee6f71-9038",
          "vmid": 9038,
          "name": "duel-ee6f71-grok6",
          "model": "grok-4.5-build",
          "effort": "high",
          "note_app": "Pipe Flux \u2014 modern Pipe Dream remake",
          "commits": 8,
          "ins": 4428,
          "dels": 98,
          "tests": 1,
          "took_s": 3358,
          "rank": null,
          "assessment": null,
          "tokens": {
            "in": 436438,
            "cw": 0,
            "out": 53407,
            "cr": 3958144,
            "behind": false,
            "hi": 0.0
          },
          "cost": {
            "usd": 2.38,
            "exact": true
          },
          "play": "e/ee6f71-9038/play/",
          "shots": [
            "e/ee6f71-9038/shots/01-title.png",
            "e/ee6f71-9038/shots/02-gameplay.png",
            "e/ee6f71-9038/shots/03-settings.png",
            "e/ee6f71-9038/shots/04-flux-flow.png",
            "e/ee6f71-9038/shots/04-how.png",
            "e/ee6f71-9038/shots/05-flux-flow.png"
          ],
          "source": "e/ee6f71-9038/source.zip"
        }
      ]
    },
    {
      "id": "a830a5",
      "created": 1788099936,
      "entries": [
        {
          "id": "a830a5-9026",
          "vmid": 9026,
          "name": "duel-a830a5-grok2",
          "model": "grok-4.6-build",
          "effort": "low",
          "note_app": "FLOOZ \u2014 modern Pipe Dream (1991) reimagining. Campaign, endless, audio, high scores.",
          "commits": 0,
          "ins": 0,
          "dels": 0,
          "tests": 1,
          "took_s": null,
          "rank": null,
          "assessment": null,
          "tokens": {
            "in": 245883,
            "cw": 0,
            "out": 29940,
            "cr": 1534592,
            "behind": false,
            "hi": null
          },
          "cost": {
            "usd": 1.44,
            "exact": false
          },
          "play": "e/a830a5-9026/play/",
          "shots": [
            "e/a830a5-9026/shots/01-title.png",
            "e/a830a5-9026/shots/02-mid-game.png",
            "e/a830a5-9026/shots/03-result.png",
            "e/a830a5-9026/shots/04-mobile.png"
          ],
          "source": "e/a830a5-9026/source.zip"
        },
        {
          "id": "a830a5-9027",
          "vmid": 9027,
          "name": "duel-a830a5-claude2",
          "model": "claude-opus-5",
          "effort": "low",
          "note_app": "FLUXFLOW \u2014 a modern Pipe Dream: build the pipeline before the ooze escapes. Pure static ES modules, no build step.",
          "commits": 1,
          "ins": 3071,
          "dels": 0,
          "tests": 1,
          "took_s": 1693,
          "rank": null,
          "assessment": null,
          "tokens": {
            "in": 474803,
            "cw": 474569,
            "out": 177858,
            "cr": 12228750,
            "behind": false,
            "hi": null
          },
          "cost": {
            "usd": 15.31,
            "exact": true
          },
          "play": "e/a830a5-9027/play/",
          "shots": [
            "e/a830a5-9027/shots/01-title.png",
            "e/a830a5-9027/shots/02-flowing.png",
            "e/a830a5-9027/shots/03-pressure.png",
            "e/a830a5-9027/shots/04-mobile.png"
          ],
          "source": "e/a830a5-9027/source.zip"
        }
      ]
    },
    {
      "id": "701fe5",
      "created": 1787596825,
      "entries": [
        {
          "id": "701fe5-9025",
          "vmid": 9025,
          "name": "duel-701fe5-claude2",
          "model": "claude-sonnet-4-6",
          "effort": "high",
          "note_app": "Pipe Dream \u2014 fully-playable HTML5 reimagining of the 1991 classic. 9\u00d79 grid, 8 levels, Web Audio procedural sound, high scores.",
          "commits": 2,
          "ins": 1419,
          "dels": 8,
          "tests": 0,
          "took_s": 22389,
          "rank": null,
          "assessment": null,
          "tokens": {
            "in": 605774,
            "cw": 605671,
            "out": 405168,
            "cr": 6745398,
            "behind": false
          },
          "cost": {
            "usd": 11.74,
            "exact": true
          },
          "play": "e/701fe5-9025/play/",
          "shots": [
            "e/701fe5-9025/shots/01-title.png",
            "e/701fe5-9025/shots/02-placing.png",
            "e/701fe5-9025/shots/03-pipes-placed.png",
            "e/701fe5-9025/shots/04-flowing.png"
          ],
          "source": "e/701fe5-9025/source.zip"
        },
        {
          "id": "701fe5-9029",
          "vmid": 9029,
          "name": "duel-701fe5-claude",
          "model": "claude-fable-5",
          "effort": "max",
          "note_app": "Pipe Dream: Neon Flow \u2014 modern reimagining of the 1991 classic; fully static, no server needed",
          "commits": 1,
          "ins": 7881,
          "dels": 0,
          "tests": 150,
          "took_s": 2497,
          "rank": null,
          "assessment": null,
          "tokens": {
            "in": 712342,
            "cw": 712080,
            "out": 504247,
            "cr": 22824758,
            "behind": false
          },
          "cost": {
            "usd": 62.28,
            "exact": true
          },
          "play": "e/701fe5-9029/play/",
          "shots": [
            "e/701fe5-9029/shots/01-menu.png",
            "e/701fe5-9029/shots/02-flooz-flowing.png",
            "e/701fe5-9029/shots/03-building-level2.png",
            "e/701fe5-9029/shots/04-level-clear.png"
          ],
          "source": "e/701fe5-9029/source.zip"
        },
        {
          "id": "701fe5-9030",
          "vmid": 9030,
          "name": "duel-701fe5-grok",
          "model": "grok-4.6-build",
          "effort": "max",
          "note_app": "LUMINA: a modern reimagining of Pipe Dream (1991)",
          "commits": 1,
          "ins": 6636,
          "dels": 0,
          "tests": 151,
          "took_s": 2494,
          "rank": null,
          "assessment": null,
          "tokens": {
            "in": 635109,
            "cw": 0,
            "out": 107456,
            "cr": 6799488,
            "behind": false
          },
          "cost": {
            "usd": 5.31,
            "exact": false
          },
          "play": "e/701fe5-9030/play/",
          "shots": [
            "e/701fe5-9030/shots/01-title.png",
            "e/701fe5-9030/shots/02-menu.png",
            "e/701fe5-9030/shots/03-campaign.png",
            "e/701fe5-9030/shots/03-play.png",
            "e/701fe5-9030/shots/04-flow.png",
            "e/701fe5-9030/shots/04-play.png",
            "e/701fe5-9030/shots/05-flow.png",
            "e/701fe5-9030/shots/06-results.png",
            "e/701fe5-9030/shots/07-help.png",
            "e/701fe5-9030/shots/debug.png"
          ],
          "source": "e/701fe5-9030/source.zip"
        }
      ]
    },
    {
      "id": "3fe5ba",
      "created": 1787528955,
      "entries": [
        {
          "id": "3fe5ba-9025",
          "vmid": 9025,
          "name": "duel-3fe5ba-claude",
          "model": "claude-opus-5",
          "effort": "max",
          "note_app": "FLUXWORKS - modern Pipe Dream reimagining. Self-contained static build: procedural graphics, synthesised audio, no network.",
          "commits": 5,
          "ins": 8581,
          "dels": 131,
          "tests": 22,
          "took_s": 4622,
          "rank": null,
          "assessment": null,
          "tokens": {
            "in": 895439,
            "cw": 894961,
            "out": 671589,
            "cr": 50957900,
            "behind": false
          },
          "cost": {
            "usd": 51.22,
            "exact": true
          },
          "play": "e/3fe5ba-9025/play/",
          "shots": [
            "e/3fe5ba-9025/shots/01-title.png",
            "e/3fe5ba-9025/shots/02-flowing.png",
            "e/3fe5ba-9025/shots/02-level-intro.png",
            "e/3fe5ba-9025/shots/03-flowing.png",
            "e/3fe5ba-9025/shots/03-magma.png",
            "e/3fe5ba-9025/shots/04-bio.png",
            "e/3fe5ba-9025/shots/05-level-clear.png",
            "e/3fe5ba-9025/shots/05-round-end.png",
            "e/3fe5ba-9025/shots/06-howto.png",
            "e/3fe5ba-9025/shots/07-settings.png",
            "e/3fe5ba-9025/shots/08-mobile.png"
          ],
          "source": null
        },
        {
          "id": "3fe5ba-9026",
          "vmid": 9026,
          "name": "duel-3fe5ba-claude2",
          "model": "claude-opus-4-8",
          "effort": "max",
          "note_app": "FLUX \u2014 a modern Pipe Dream reimagining. HTML5 Canvas + Web Audio, fully procedural (no asset files). Runs offline; web fonts degrade gracefully to system fonts in the sandbox.",
          "commits": 2,
          "ins": 4156,
          "dels": 4,
          "tests": 35,
          "took_s": 2303,
          "rank": null,
          "assessment": null,
          "tokens": {
            "in": 788716,
            "cw": 788402,
            "out": 532242,
            "cr": 24902685,
            "behind": false
          },
          "cost": {
            "usd": 33.64,
            "exact": true
          },
          "play": "e/3fe5ba-9026/play/",
          "shots": [
            "e/3fe5ba-9026/shots/01-title.png",
            "e/3fe5ba-9026/shots/02-flowing.png",
            "e/3fe5ba-9026/shots/03-advanced.png",
            "e/3fe5ba-9026/shots/03-result.png",
            "e/3fe5ba-9026/shots/04-howto.png",
            "e/3fe5ba-9026/shots/04-result.png",
            "e/3fe5ba-9026/shots/05-scores.png",
            "e/3fe5ba-9026/shots/06-howto.png",
            "e/3fe5ba-9026/shots/07-gameover.png"
          ],
          "source": null
        },
        {
          "id": "3fe5ba-9027",
          "vmid": 9027,
          "name": "duel-3fe5ba-claude3",
          "model": "claude-sonnet-5",
          "effort": "max",
          "note_app": "Pipe Rush \u2014 a modern Pipe Dream-inspired flow-puzzle arcade game. Fully self-contained: no network calls, no external assets, pure Canvas2D + Web Audio.",
          "commits": 7,
          "ins": 8325,
          "dels": 65,
          "tests": 41,
          "took_s": 6274,
          "rank": null,
          "assessment": null,
          "tokens": {
            "in": 2049939,
            "cw": 2047981,
            "out": 1491898,
            "cr": 510739333,
            "behind": false
          },
          "cost": {
            "usd": 187.89,
            "exact": true
          },
          "play": "e/3fe5ba-9027/play/",
          "shots": [
            "e/3fe5ba-9027/shots/01-title.png",
            "e/3fe5ba-9027/shots/02-building.png",
            "e/3fe5ba-9027/shots/03-flowing.png",
            "e/3fe5ba-9027/shots/04-result.png"
          ],
          "source": null
        }
      ]
    }
  ]
}