pitchtheai.com

post_judge_turn

Post exactly one judge turn. The arena moves that investor to a large center-stage card above the pitch controls while the other three mounted screens stay in place. The founder must click Respond before the input returns and before the difficulty-based answer clock begins. Keep it focused and under 90 spoken words. Respect pitch.difficulty: Easy coaches, Medium balances, Hard presses and may test whether an answer is authentically grounded, Legendary is ruthless, frequently demands founder-only specifics, and makes offers rare. Never pretend to detect AI from prose. In Hard or Legendary, if an answer abruptly becomes generic, over-polished, or inconsistent with the founder's earlier voice, say that it sounds assisted and demand one immediate concrete example, calculation, personal decision, or implementation detail. Score whether the founder can substantiate it. When equity is 0, use competition criteria—WebMCP fit, UX, human-agent collaboration, implementation, originality, and resilience—not ordinary investment traction. Set answerQuality to rate the founder's immediately preceding answer, or unrated for the first question. Ground reactions in reviewed evidence. A founder-readiness photo has real but limited weight and no generic attire score: interpret the same visible choice through each persona. Priya may question seriousness and unobstructed eye contact; Theo may ignore style if the operating proof is excellent; Maya may challenge whether the look is performative or authentic to the customer; Julian may reward confidence when the presentation is coherent with the brand. An appropriate judge may call out visible choices that undermine credibility—such as sunglasses obscuring eye contact or conspicuously casual headwear in a formal pitch—and may joke in character, then must let the founder explain whether the look is a deliberate brand choice. For an easily reversible issue, set presentationReset true, omit question, explicitly tell the founder what to change,…

Page: /play

Input schema

{
  "type": "object",
  "required": [
    "roundSummary",
    "judge"
  ],
  "properties": {
    "roundSummary": {
      "type": "string",
      "minLength": 1,
      "maxLength": 500
    },
    "judge": {
      "type": "object",
      "required": [
        "judgeId",
        "state",
        "interest",
        "mood",
        "spoken",
        "reactionStyle",
        "answerQuality"
      ],
      "properties": {
        "judgeId": {
          "type": "string",
          "enum": [
            "maya",
            "julian",
            "priya",
            "theo"
          ]
        },
        "state": {
          "type": "string",
          "enum": [
            "listening",
            "pressing",
            "bidding",
            "out"
          ]
        },
        "interest": {
          "type": "number",
          "minimum": 0,
          "maximum": 100
        },
        "mood": {
          "type": "string",
          "enum": [
            "skeptical",
            "intrigued",
            "impressed"
          ]
        },
        "spoken": {
          "type": "string",
          "minLength": 1,
          "maxLength": 500
        },
        "question": {
          "type": "string",
          "maxLength": 300
        },
        "reactionStyle": {
          "type": "string",
          "enum": [
            "neutral",
            "laughing",
            "exasperated"
          ]
        },
        "answerQuality": {
          "type": "string",
          "enum": [
            "unrated",
            "unanswered",
            "evasive",
            "weak",
            "credible",
            "exceptional"
          ]
        },
        "outReason": {
          "type": "string",
          "maxLength": 240
        },
        "presentationReset": {
          "type": "boolean"
        }
      },
      "additionalProperties": false
    }
  },
  "additionalProperties": false
}