business-frameworks.matryoshka-paradigms.workers.dev

Live 402 challenge

Captured by the enrichment pass, not read just now. Prices can change — always read the 402 the endpoint answers with.

{
  "fit": {
    "arms": {
      "without": "the model alone, no skill, no node",
      "with_node": "the skill plus the paid node text for the question's decision",
      "free_layer": "the skill installed and told to use it: the free layer and the handles, no paid text"
    },
    "kind": "screen",
    "pilot": true,
    "not_for": [
      "market data, prices or forecasts — the nodes carry none",
      "businesses above about $50M in revenue, or questions outside value, cash flow, risk and growth",
      "arithmetic over periods, or an order of steps answered from the text alone — the applied route computes them; the text does not",
      "a frontier model already on a screen-or-threshold question — no lift was established at this sample size"
    ],
    "verdict": {
      "because": "This model was not measured; no figure is given for it. The two measured rows are the only evidence; a model below the frontier class is more likely to behave like the small-model row.",
      "verdict": "not_measured",
      "for_model": null,
      "model_class": "unknown"
    },
    "by_model": {
      "claude-opus-5.5": {
        "class": "frontier",
        "because": "On screens and thresholds, a frontier model stated the rule in 4 of 5 answers with the node against 1 of 5 alone — too few answers to call that a lift (across all kinds, 5 of 8 against 2 of 8, p = 0.31), and an answer with the node costs about 1.7× an unaided one at list prices (4.8× the raw tokens, most of them cache reads). If you want a computed verdict, the applied route gives one; otherwise nothing here shows your own answer to be worse.",
        "verdict": "unproven",
        "this_decision_alone": {
          "n": 2,
          "note": "too few answers to read on their own; the kind row is the evidence",
          "rule_right": {
            "without": 0,
            "with_node": 2,
            "free_layer": 2
          }
        },
        "this_kind_of_question": {
          "n": 5,
          "rule_right": {
            "without": 1,
            "with_node": 4,
            "free_layer": 3
          },
          "tokens_per_answer": {
            "without": {
              "output": 1256,
              "input_new": 7543,
              "input_total": 14974
            },
            "with_node": {
              "output": 1540,
              "input_new": 12638,
              "input_total": 76535
            },
            "free_layer": {
              "output": 1929,
              "input_new": 12811,
              "input_total": 80462
            }
          },
          "cost_per_answer_usd": {
            "without": 0.08695,
            "with_node": 0.14465,
            "free_layer": 0.15456
          },
          "api_calls_per_answer": {
            "without": 1,
            "with_node": 4,
            "free_layer": 4.2
          },
          "cost_per_rule_right_answer_usd": {
            "without": 0.43475,
            "with_node": 0.18081,
            "free_layer": 0.2576
          }
        },
        "list_price_usd_per_mtok": {
          "input": 4,
          "output": 20
        }
      },
      "claude-haiku-5.5": {
        "class": "small",
        "because": "On screens and thresholds, a small model stated the rule in 12 of 12 answers with the node, 7 of 12 with the free layer alone and 2 of 12 with nothing. An answer with the node costs about 2.0× an unaided one at list prices (3.8× the raw tokens, most of them cache reads), and still costs less per rule-right answer than any other arm measured.",
        "verdict": "use",
        "this_decision_alone": {
          "n": 2,
          "note": "too few answers to read on their own; the kind row is the evidence",
          "rule_right": {
            "without": 0,
            "with_node": 2,
            "free_layer": 2
          }
        },
        "this_kind_of_question": {
          "n": 12,
          "rule_right": {
            "without": 2,
            "with_node": 12,
            "free_layer": 7
          },
          "tokens_per_answer": {
            "without": {
              "output": 1507,
              "input_new": 7333,
              "input_total": 18894
            },
            "with_node": {
              "output": 2579,
              "input_new": 14107,
              "input_total": 75373
            },
            "free_layer": {
              "output": 2368,
              "input_new": 12791,
              "input_total": 87758
            }
          },
          "cost_per_answer_usd": {
            "without": 0.00234,
            "with_node": 0.00472,
            "free_layer": 0.00449
          },
          "api_calls_per_answer": {
            "without": 1.17,
            "with_node": 3.67,
            "free_layer": 4.33
          },
          "cost_per_rule_right_answer_usd": {
            "without": 0.01404,
            "with_node": 0.00472,
            "free_layer": 0.0077
          }
        },
        "list_price_usd_per_mtok": {
          "input": 0.1,
          "output": 0.5
        }
      }
    },
    "decision": "risk/decide/concentration",
    "table_url": "https://business-frameworks.matryoshka-paradigms.workers.dev/fit",
    "kind_label": "screens and thresholds",
    "measured_on": "2026-10-08",
    "will_change": "the comparison test (three runs per question, a second vendor, about sixty runs per arm) replaces these figures; measured_on changes when it does",
    "your_own_cost": "multiply tokens_per_answer by your model's list price: input_total counts every input token across the calls one answer took (most of it cache reads on the second and later calls), input_new the uncached part, output the tokens written; the measured cost_per_answer_usd used the prices in models.list_price_usd_per_mtok",
    "recommended_path": {
      "path": "a small model with the node",
      "what": "pin a small model for these questions and buy the node for each: the measured rule-right rate is the highest of any arm, at the lowest cost per rule-right answer",
      "quality": "quality here is rule-right: the answer states this product's rule for the question, graded by a model against keys written from the nodes — a pilot of one run per question, not a customer-satisfaction measure (a blind-preference arm is planned)",
      "the_delta": "per rule-right answer on screens and thresholds: $0.0047 for a small model with the node against $0.435 for a frontier model alone (92×) and $0.181 for a frontier model with the node (38×); across all kinds $0.0064 against $0.335 (52×) and $0.241 (37×)",
      "rule_right": {
        "across_all_kinds": {
          "note": "small model with the node against frontier with the node: no difference detected (p = 0.65)",
          "frontier_model_alone": "2 of 8",
          "small_model_with_node": "15 of 20",
          "frontier_model_with_node": "5 of 8"
        },
        "frontier_model_alone": "1 of 5",
        "small_model_with_node": "12 of 12",
        "frontier_model_with_node": "4 of 5"
      },
      "cost_per_rule_right_answer_usd": {
        "across_all_kinds": {
          "frontier_model_alone": 0.3352,
          "small_model_with_node": 0.00644,
          "frontier_model_with_node": 0.2412
        },
        "frontier_model_alone": 0.43475,
        "small_model_with_node": 0.00472,
        "frontier_model_with_node": 0.18081
      },
      "if_you_cannot_choose_the_model": "on a frontier model already: the node's lift is unproven at this sample size (see the verdict); on a small model: use the node; on a model not measured: no figure — the small-model row is the nearest evidence, not a promise"
    },
    "runs_per_question": 1,
    "ask_for_your_model": "https://business-frameworks.matryoshka-paradigms.workers.dev/fit?decision=risk/decide/concentration&model=<your model name>",
    "list_prices_read_on": "2026-10-09",
    "what_rule_right_means": "the answer states the decision rule of the node that covers the question, as graded by a model against keys written from the nodes; it is agreement with this product's rule, not an outside standard of correctness"
  },
  "kind": "node",
  "error": "payment_required",
  "layer": 3,
  "title": "Limit concentration",
  "access": "Ask narrow, pay per node; ask broad, get a quoted bundle whose price is the sum of the nodes it takes, stated before you pay.",
  "handle": "business-frameworks/risk/decide/concentration@1.2",
  "tokens": 137,
  "cite_as": "McHenry, J. (2026). Business Frameworks, version 1.2, business-frameworks/risk/decide/concentration@1.2.",
  "payment": {
    "asset": "USDC",
    "pay_to": "0xbBaCd5111E0Ca3a52DdAA39D4723Cf75C1A75D91",
    "scheme": "exact",
    "network": "eip155:8453",
    "protocol": "x402",
    "terms_header": "PAYMENT-REQUIRED",
    "x402_version": 2,
    "amount_atomic": "50000",
    "asset_contract": "0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913",
    "max_timeout_seconds": 300
  },
  "price_usd": "$0.05",
  "how_to_pay": {
    "limits": "Pay only pay_to, only USDC on Base, never more than price_usd, once per answer. A failed or malformed request is not charged.",
    "as_a_person": "Open this URL in a browser that has a wallet holding USDC on Base; the page that loads offers the payment.",
    "without_a_wallet": "Do not pay. Tell the user the price and what it buys. The free layer stays open at catalogue_url.",
    "with_a_wallet_tool": {
      "call": {
        "tool": "initiate_x402_request",
        "arguments": {
          "url": "https://business-frameworks.matryoshka-paradigms.workers.dev/node/risk/decide/concentration",
          "method": "GET",
          "maxPayment": "0.05"
        }
      },
      "note": "If your harness has a wallet tool that speaks x402, call it with this URL and a cap equal to price_usd. Example, Coinbase Wallet MCP:",
      "then": "The user approves in their wallet; then call complete_x402_request with the returned requestId."
    },
    "with_an_x402_client": "Repeat this request with a PAYMENT-SIGNATURE header built from the PAYMENT-REQUIRED header on this response. Set the client's spend cap to price_usd."
  },
  "catalogue_url": "https://business-frameworks.matryoshka-paradigms.workers.dev/catalogue"
}

Accepts

The payment requirements as published to the catalog. Read the live 402 before paying — a price here is a claim, not a quote.

[
  {
    "asset": "0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913",
    "extra": {
      "name": "USD Coin",
      "version": "2"
    },
    "payTo": "0xbBaCd5111E0Ca3a52DdAA39D4723Cf75C1A75D91",
    "scheme": "exact",
    "network": "eip155:8453",
    "mimeType": "application/json",
    "resource": "https://business-frameworks.matryoshka-paradigms.workers.dev/node/risk/decide/concentration",
    "description": "",
    "maxAmountRequired": "50000",
    "maxTimeoutSeconds": 300
  }
]

Provenance

Seen in the source catalog

2026-10-07 05:57Z

Last indexed by Roundhouse

2026-10-26 04:45Z

Last enriched (probe, favicon, geo)

2026-10-11 15:16Z

x402 version

1

Max timeout

300s

Liveness probe

HTTP 402