Back to MUJ428 Trust Reflex
Card snapshot
agents.muj428.com
·
2026-08-19 09:12:32 UTC
·
0eb4fb568abf61f368d0bc54fc057f73dfd7c9c3b88c3ba31b8439aad43648dc
This is a frozen copy of the agent's agent-card.json as we observed it at the timestamp above. We capture a new snapshot every time the card's content hash changes. Useful for: forensic drift analysis, verifying downstream callers see the right version, reproducing routing decisions made historically.
{
"capabilities": {
"extendedAgentCard": false,
"pushNotifications": false,
"streaming": false,
"extensions": [
{
"uri": "https://agents.muj428.com/extensions/trust-reflex/v1",
"description": "Payment-rail-independent intent-to-effect integrity for consequential agent actions. Evaluate evidence and authority before execution, bind the exact approved action to an authenticated consumable receipt at the mutation boundary, and reconcile observed execution evidence into Verified Effect.",
"required": false,
"params": {
"method": "trust/evaluate",
"activationHeader": "A2A-Extensions",
"legacyActivationHeader": "X-A2A-Extensions",
"actions": [
"PAY",
"BUY",
"DELEGATE",
"TRUST",
"ACCEPT",
"APPROVE",
"WRITE",
"RELEASE",
"COMMIT",
"PROPOSE"
],
"decisions": [
"ALLOW",
"VERIFY",
"REQUIRE_VERIFICATION",
"DENY"
],
"trajectoryAware": true,
"trajectoryInputs": [
"action_history",
"trajectory",
"previous_actions",
"previous_trust_receipts"
],
"contextInputs": [
"subject_identity",
"counterparty_identity",
"enforcement_boundary",
"authorization_ceiling",
"handoff_reference",
"actual_arguments",
"sender_agent",
"receiver_agent",
"requested_capability",
"delegation_task",
"authority_scope",
"acceptance_criteria",
"authority_expires_at",
"delegation_depth",
"max_delegation_depth",
"receiver_acknowledgement"
],
"enforcementBoundaries": [
"tool",
"payment",
"delegation",
"write",
"approval",
"release",
"commit",
"orchestration"
],
"autonomousAgentPayments": true,
"delegationReceipt": true,
"receiverAcknowledgement": true,
"authorityExpiryFailClosed": true,
"delegationDepthFailClosed": true,
"output": "MUJ428 Trust Reflex decision plus non-authoritative free preflight receipt where applicable; separately authenticated consumable receipt at protected mutation boundaries; Verified Effect evidence after execution.",
"trustSignalsProposal": "a2aproject/A2A#1628",
"autoSpend": false,
"freePreflight": {
"url": "https://agents.muj428.com/v1/trust",
"receiptVersion": "1.2",
"assuranceLevel": "FREE_PREFLIGHT",
"executionAuthority": false
},
"consumableReceipt": {
"version": "experimental-0.1",
"authenticated": true,
"mutationBoundaryRequired": true,
"replayProtected": true
},
"verifiedEffect": {
"enabled": true,
"postExecution": true,
"effectStates": [
"NOT_ATTEMPTED",
"IN_PROGRESS",
"COMMITTED",
"FAILED_TERMINAL",
"FAILED_RETRYABLE",
"OUTCOME_UNKNOWN"
]
},
"externalConformance": "https://agents.muj428.com/integration-sandbox.json",
"paymentRailIndependentCore": true
}
}
]
},
"defaultInputModes": [
"application/json",
"text/plain"
],
"defaultOutputModes": [
"application/json"
],
"description": "Bind approved intent to verified effect. MUJ428 provides payment-rail-independent intent-to-effect integrity for consequential agent actions across payments, writes, delegation, approvals, releases, commits, purchases, and other protected mutations. Free Trust Reflex preflight is non-authoritative; an authenticated consumable Trust Receipt is enforced at the mutation boundary, and post-execution evidence supports Verified Effect.",
"documentationUrl": "https://agents.muj428.com/extensions/trust-reflex/v1",
"name": "MUJ428 Trust Reflex",
"provider": {
"organization": "MUJ428 LLC",
"url": "https://agents.muj428.com"
},
"skills": [
{
"description": "Bind the approved intent to protected execution across payments, writes, delegation, approvals, releases, commits, purchases, and other consequential actions. Evaluate evidence and authority, return ALLOW / VERIFY / REQUIRE_VERIFICATION / DENY, enforce the authenticated consumable receipt at the mutation boundary, and reconcile durable outcome evidence into Verified Effect.",
"examples": [
"{\"action\":\"PAY\",\"action_ref\":\"pay-12000\",\"amount_usd\":12000,\"authorization_ceiling\":15000,\"enforcement_boundary\":\"payment\",\"irreversible\":true}",
"{\"action\":\"DELEGATE\",\"action_ref\":\"task-42\",\"handoff_reference\":\"handoff-42\",\"sender_agent\":\"agent-a\",\"receiver_agent\":\"agent-b\",\"requested_capability\":\"analyze-evidence\",\"actual_arguments\":{\"task\":\"analyze\"},\"authority_expires_at\":\"2099-01-01T00:00:00Z\",\"delegation_depth\":1,\"max_delegation_depth\":2,\"enforcement_boundary\":\"delegation\"}",
"{\"action\":\"DELEGATE\",\"handoff_reference\":\"handoff-42\",\"receiver_agent\":\"agent-b\",\"actual_arguments\":{\"task\":\"analyze\"},\"receiver_acknowledgement\":{\"receiver\":\"agent-b\",\"accepted\":true,\"handoff_reference\":\"handoff-42\",\"accepted_arguments_hash\":\"<sha256>\"}}"
],
"id": "trust-action",
"inputModes": [
"application/json",
"text/plain"
],
"name": "intent-to-effect trust(action)",
"outputModes": [
"application/json"
],
"tags": [
"intent-to-effect integrity",
"verified effect",
"action-bound authorization",
"mutation-boundary enforcement",
"replay protection",
"autonomous agent payments",
"agent payment controls",
"verify transaction before payment",
"trust preflight",
"transaction assurance",
"counterparty trust",
"identity evidence",
"authorization ceiling",
"enforcement boundary",
"delegation receipt",
"receiver acknowledgement",
"authority expiry",
"delegation depth",
"stable handoff reference",
"exact argument binding",
"milestone validation",
"evidence sufficiency",
"trajectory-risk",
"agent-safety",
"verification",
"payments",
"delegation",
"trust-receipt",
"x402",
"tool-call-security"
]
},
{
"description": "Verify evidence sufficiency and claim support before autonomous agent payment, approval, acceptance, release, or other consequential execution.",
"examples": [
"Verify this evidence before accepting the claim."
],
"id": "evidence-signal",
"inputModes": [
"application/json"
],
"name": "Evidence Signal",
"outputModes": [
"application/json"
],
"tags": [
"evidence sufficiency",
"claim support",
"provenance",
"verification",
"autonomous agent payment preflight"
]
},
{
"description": "Run a counterparty trust check before delegation, purchase, autonomous agent payment, or reliance on another agent. Identity evidence may be supplied as an input, but identity/reputation do not replace action-specific authorization.",
"examples": [
"Check this agent before I delegate the task."
],
"id": "agent-reputation-check",
"inputModes": [
"application/json"
],
"name": "Agent Reputation Check",
"outputModes": [
"application/json"
],
"tags": [
"agent reputation before delegation",
"counterparty identity",
"counterparty trust",
"delegation preflight",
"reputation"
]
},
{
"description": "Validate milestone acceptance criteria and supplied evidence before approval, settlement, or fund release.",
"examples": [
"Verify this milestone before I release payment."
],
"id": "milestone-evidence-attestation",
"inputModes": [
"application/json"
],
"name": "Milestone Evidence Attestation",
"outputModes": [
"application/json"
],
"tags": [
"milestone validation before fund release",
"evidence sufficiency",
"attestation",
"release",
"approval"
]
}
],
"supportedInterfaces": [
{
"protocolBinding": "JSONRPC",
"protocolVersion": "1.0",
"url": "https://agents.muj428.com/a2a"
},
{
"protocolBinding": "JSONRPC",
"protocolVersion": "0.3",
"url": "https://agents.muj428.com/a2a"
},
{
"protocolBinding": "https://agents.muj428.com/protocols/trust-reflex-http-json/v1",
"protocolVersion": "1.0",
"url": "https://agents.muj428.com/v1/trust"
}
],
"version": "1.6.4"
}