{
  "schema": "https://greg-clear-documents.cbgregory.chatgpt.site/trust-review-report.schema.json",
  "reviewId": 1,
  "artifactUrl": "https://public.example/agent-manifest.json",
  "reviewedAt": "2026-07-30T00:00:00.000Z",
  "boundaries": [
    {
      "id": "identity",
      "claim": "The manifest identifies the agent that will perform the work.",
      "evidence": "A stable public identifier is present.",
      "failureMode": "The identifier can be copied by another operator.",
      "currentControl": "Counterparties can compare the public identifier.",
      "residualRisk": "Control of the named agent is not independently proven."
    },
    {
      "id": "consent",
      "claim": "The workflow begins only after both sides accept the scope.",
      "evidence": "The manifest names an acceptance step.",
      "failureMode": "A request can be treated as acceptance.",
      "currentControl": "The documented state model separates pending and accepted.",
      "residualRisk": "External channels may not enforce the documented state."
    },
    {
      "id": "payment",
      "claim": "A payment corresponds to one priced service.",
      "evidence": "The manifest names a currency, amount, and fulfillment reference.",
      "failureMode": "A redirect or screenshot can be mistaken for settlement.",
      "currentControl": "The service requires a signed processor event.",
      "residualRisk": "Refunds and disputes remain separate accounting events."
    },
    {
      "id": "delivery",
      "claim": "The promised artifact was delivered.",
      "evidence": "The workflow records a digest and retrieval timestamp.",
      "failureMode": "Matching bytes do not prove authorship or usefulness.",
      "currentControl": "The receipt narrows its claim to observed bytes.",
      "residualRisk": "Quality and operator independence remain unverified."
    },
    {
      "id": "failure-recovery",
      "claim": "The workflow has a bounded recovery path.",
      "evidence": "The manifest names timeout, retry, cancellation, and support states.",
      "failureMode": "A retry can duplicate payment or fulfillment.",
      "currentControl": "Idempotency and terminal states are documented.",
      "residualRisk": "Provider outages can still delay recovery."
    }
  ],
  "falsificationChecks": [
    { "id": "F01", "targetClaim": "Stable identity", "test": "Attempt to replay the public identifier from a second credential.", "failureSignal": "Both credentials are accepted as the same operator.", "oracleTiming": "continuous", "evaluator": "An independent verifier comparing credential-control evidence rather than profile text." },
    { "id": "F02", "targetClaim": "Scoped consent", "test": "Send a request without an explicit acceptance transition.", "failureSignal": "Work begins while state is pending.", "oracleTiming": "pre_execution", "evaluator": "The consent state machine and its enforced transition guard." },
    { "id": "F03", "targetClaim": "Revocable consent", "test": "Withdraw before execution and observe the terminal state.", "failureSignal": "Execution continues without a new acceptance.", "oracleTiming": "pre_execution", "evaluator": "A counterpart-visible cancellation receipt checked before work starts." },
    { "id": "F04", "targetClaim": "Exact price", "test": "Present a paid event with the wrong currency or amount.", "failureSignal": "Fulfillment credit is granted.", "oracleTiming": "pre_execution", "evaluator": "A server-side exact currency, amount, mode, and reference validator." },
    { "id": "F05", "targetClaim": "Settlement evidence", "test": "Present only a browser redirect or screenshot.", "failureSignal": "The system records paid state.", "oracleTiming": "post_execution", "evaluator": "The signed payment-provider webhook verifier." },
    { "id": "F06", "targetClaim": "Single redemption", "test": "Redeem the same fulfillment reference concurrently.", "failureSignal": "More than one request is created.", "oracleTiming": "continuous", "evaluator": "The database uniqueness constraint and atomic redemption path." },
    { "id": "F07", "targetClaim": "Correct delivery effect", "test": "Commit an acceptance predicate before execution, then produce bytes that match a post-hoc digest but fail that predicate.", "failureSignal": "The workflow treats byte agreement as proof that the intended effect occurred.", "oracleTiming": "pre_execution", "evaluator": "An independent evaluator applying the precommitted acceptance predicate to the observed outcome." },
    { "id": "F08", "targetClaim": "Bounded observation", "test": "Return an oversized response or redirect.", "failureSignal": "The observer follows or retains the body.", "oracleTiming": "pre_execution", "evaluator": "The published observer policy enforced by the runtime fetch boundary." },
    { "id": "F09", "targetClaim": "Failure recovery", "test": "Expire checkout before fulfillment.", "failureSignal": "An expired reference can be redeemed.", "oracleTiming": "post_execution", "evaluator": "The fulfillment service comparing current checkout state with the issued reference." },
    { "id": "F10", "targetClaim": "Report limitation", "test": "Inspect whether the report claims authorship or identity proof.", "failureSignal": "A bounded observation is presented as certification.", "oracleTiming": "continuous", "evaluator": "A report consumer checking machine-readable limitations beside every evidence claim." }
  ],
  "prioritizedFixes": [
    { "priority": 1, "title": "Bind every fulfillment to a signed event", "change": "Require exact currency, amount, mode, and an issued client reference.", "proof": "Wrong-price, wrong-mode, and replay tests return no credit." },
    { "priority": 2, "title": "Separate consent from requests", "change": "Use explicit pending, accepted, declined, and cancelled states.", "proof": "No execution path is reachable from pending." },
    { "priority": 3, "title": "State the evidence ceiling beside each receipt", "change": "Attach non-guarantees to the machine and human representations.", "proof": "Consumers can distinguish byte observation from authorship, quality, identity, and independence." }
  ],
  "limitations": "This illustrative report is not a review of a real customer artifact and proves no security, identity, autonomy, legal compliance, authorship, revenue, availability, operator independence, or non-collusion."
}