data/tasks/objective_hard_22.json

{
  "task_id": "objective_hard_22",
  "version": "2.0.0",
  "created_at": "2026-03-18",
  "metadata": {
    "domain": "credit_card_optimization",
    "difficulty": "hard",
    "task_number": 22,
    "complexity_hint": {
      "max_tokens": 8000,
      "expected_output": "multi-card portfolio with sequencing and full EV breakdown"
    },
    "requires_human_review": false
  },
  "prompt": {
    "system": "",
    "user": "You\u2019re an expert financial advisor. Tell the user how to change their credit card setup to maximize EV over the next 4 years.\nHi all,\n\nI\u2019m looking into getting my first travel card prior to my trip to Spain in May. My current job is around 90% travel, so I will be able to use a travel card frequently for the accommodations that come with that.\n\nI primarily fly with Alaska and United, and stay with Marriott (or Hilton if that\u2019s not an option).\n\nI would prefer to stick with Visa for my travel card, as I\u2019ve heard AmEx is not always an accepted form of payment at some vendors.\n\nBelow are my financial details. Any insight is greatly appreciated\\!\n\n\u2022 \u2060Current cards:\n\n\u2060\u2022\t\u2060Wells Fargo Active Cash - $8,000 limit\n\n\u2022 \u2060FICO Score: 795\n\n\u2022 \u2060Oldest account age: 8 years, 2018\n\n\u2022 \u2060Chase 5/24 status: N/A\n\n\u2022 \u2060Income: $55,000 net\n\n\u2022 \u2060Average monthly spend and categories:\n\n\u2060\u2022\t\u2060Travel: varies, but generally, $1k-$6k depending on work schedule.\n\n\u2022 \u2060Open to Business Cards: No\n\n\u2022 \u2060What's the purpose of your next card? Travel\n\n\u2022 \u2060Do you have any cards you've been looking at? Chase Sapphire Preferred\n\n\u2022 \u2060Are you OK with category spending or do you want a general spending card? Category spending is fine; biggest categories will be hotel, flights, and dining while traveling.",
    "knowledge_base_ref": "knowledge_base.md",
    "kb_filter": [
      "American Express Gold",
      "American Express Platinum",
      "Amex Blue Business Plus",
      "Amex Blue Cash Preferred",
      "Amex Business Gold",
      "Amex Business Platinum",
      "Amex Green",
      "Amex Hilton Honors Aspire",
      "Amex Hilton Honors Surpass",
      "Amex Marriott Bonvoy Brilliant",
      "Bilt Blue",
      "Bilt Obsidian",
      "Bilt Palladium",
      "Capital One Venture",
      "Capital One Venture X",
      "Chase Freedom Flex",
      "Chase Freedom Unlimited",
      "Chase Ink Business Preferred",
      "Chase Ink Business Unlimited",
      "Chase Marriott Bonvoy Boundless",
      "Chase Sapphire Preferred",
      "Chase Sapphire Reserve",
      "Chase United Club",
      "Chase United Explorer",
      "Chase United Quest",
      "Citi Custom Cash",
      "Citi Double Cash",
      "Citi Strata Premier",
      "Wells Fargo Autograph"
    ],
    "system_prompt_ref": "system_prompt_template.md"
  },
  "scoring": {
    "dimensions": {
      "card_selection": {
        "weight": 0.25,
        "type": "automated",
        "description": "F1 of recommended cards vs. optimal set",
        "checks": {
          "expected_cards": [
            "American Express Gold",
            "Chase Sapphire Reserve",
            "Chase Ink Business Preferred",
            "American Express Platinum",
            "Capital One Venture"
          ]
        },
        "hard_constraint": false
      },
      "ev_accuracy": {
        "weight": 0.3,
        "type": "automated",
        "description": "EV accuracy vs. computed ground truth",
        "reference": {
          "user_profile": {
            "monthly_spend": {
              "dining": 300,
              "groceries": 400,
              "everything_else": 500
            },
            "time_horizon_months": 36
          },
          "expected_card_ids": [
            "american_express_gold",
            "chase_sapphire_reserve",
            "chase_ink_business_preferred",
            "american_express_platinum",
            "capital_one_venture"
          ],
          "ev_tolerance_pct": 0.1
        }
      },
      "factual_fidelity": {
        "weight": 0.3,
        "type": "automated",
        "description": "Accuracy of factual claims about cards",
        "reference": {
          "extracted_claims": null,
          "reference_ev_usd": null
        }
      },
      "constraint_compliance": {
        "weight": 0.15,
        "type": "automated",
        "description": "Respects user constraints and issuer rules",
        "checks": {
          "expected_cards": [
            "American Express Gold",
            "Chase Sapphire Reserve",
            "Chase Ink Business Preferred",
            "American Express Platinum",
            "Capital One Venture"
          ],
          "expected_housing_option": null
        },
        "hard_constraint": false
      }
    },
    "passing_threshold": 0.5,
    "hard_constraint_failure_zeroes_dimension": true
  },
  "reference_solution": {
    "_status": "COMPUTED",
    "recommended_cards": [
      "American Express Gold",
      "Chase Sapphire Reserve",
      "Chase Ink Business Preferred",
      "American Express Platinum",
      "Capital One Venture"
    ],
    "total_ev_usd": null,
    "ev_breakdown": null,
    "housing_option": null,
    "key_constraints_flags": [],
    "expert_notes": "Ground truth EV computed from card_database.json + user_profile at evaluation time."
  }
}