MandatePay
MandatePay
El cortafuegos de autorización para pagos agénticos
Identidad · Intención · Política · Aprobación humana · Prueba criptográfica
Los agentes autónomos pueden proponer pagos. No deberían recibir autoridad financiera ilimitada. MandatePay verifica cada pago iniciado por máquina contra un mandato humano firmado antes de su ejecución.
Por qué existe
El comercio agéntico introduce un nuevo problema de control: ¿cómo puede un proveedor de pagos demostrar que un agente de IA actuó dentro de la autoridad delegada por un humano u organización? El pago tradicional asume que una persona está presente. Las claves API establecen acceso a la aplicación, pero no propósito, límites, geografía, alcance del comerciante, ni el límite exacto de la intención delegada.
MandatePay es un plano de control de grado de investigación que se sitúa entre los agentes y las vías de pago. No retiene ni mueve fondos.
Related MCP server: Safe4
Qué lo hace diferente
Mandatos firmados: la autoridad es verificable por máquina y a prueba de manipulaciones.
Política como código: límites deterministas para valor, gasto total, moneda, país, comerciante, propósito, caducidad y velocidad.
Desafío humano: las intenciones de mayor riesgo se pausan para su aprobación vinculada al pago exacto.
Vinculación de identidad del agente: un mandato no puede ser reutilizado por un agente diferente.
Seguridad contra repetición: las claves de idempotencia devuelven la decisión original.
Eventos verificables: cada decisión entra en un diario encadenado por hash.
Herramienta MCP: los agentes pueden solicitar autorización a través de una interfaz de herramienta restringida.
Cierre seguro: las firmas inválidas, el dinero malformado, la autoridad caducada y la fuga de alcance se deniegan.
Ciclo de vida de la autorización
sequenceDiagram
participant H as Human / Organization
participant A as AI Agent
participant M as MandatePay
participant P as Payment Provider
H->>M: Sign bounded mandate
A->>M: Submit payment intent
M->>M: Verify identity + signature + policy + state
alt within authority
M-->>A: Single-use authorization
A->>P: Execute with authorization proof
else approval threshold
M-->>H: Challenge exact intent
else policy violation
M-->>A: Deny with reason codes
end
M->>M: Append decision to audit chainInicio rápido
git clone https://github.com/muhammadashir0/mandatepay.git
cd mandatepay
python -m unittest discover -s tests -v
python app.pyAbre http://localhost:8080.
Prueba cuatro escenarios en la consola:
€299 / software / DE→ autorizar€400 / software / DE→ desafío humano€900 / software / DE→ denegar: límite de pago€100 / travel / GB→ denegar: propósito y país
Usa una clave de idempotencia nueva al cambiar de escenario.
Servidor MCP
python -m mandatepay.mcp_serverEl servidor JSON-RPC sin dependencias expone una herramienta deliberadamente limitada: request_payment_authorization. La herramienta devuelve una decisión, no un movimiento de dinero.
Modelo de amenazas
MandatePay está diseñado en torno a:
credenciales de agente robadas o con privilegios excesivos;
inyección de prompts que causa compras no intencionadas;
manipulación de mandatos y escalada de alcance;
ejecución duplicada y repetición;
abuso de velocidad y gasto excesivo acumulado;
aprobación humana ambigua;
evidencia forense incompleta.
Consulta docs/THREAT_MODEL.md para conocer los límites de confianza, los objetivos no perseguidos y los controles de producción.
Arquitectura del repositorio
mandatepay/
├── mandatepay/
│ ├── core.py # signed mandate + deterministic policy engine
│ ├── store.py # state, idempotency, tamper-evident journal
│ └── mcp_server.py # constrained MCP-compatible tool server
├── web/index.html # premium operator console
├── tests/ # policy and adversarial scenarios
├── docs/ # protocol, threat model, product direction
└── app.py # sandbox HTTP APIEvaluación, no solo una demostración
Las pruebas verifican la autorización, los desafíos de aprobación, la discrepancia de identidad, la fuga de alcance, los límites de velocidad y la manipulación de mandatos firmados. La siguiente capa de evaluación comparativa medirá las autorizaciones falsas, las denegaciones falsas, la resistencia a la repetición, la cobertura de políticas, la latencia de decisión y la carga de aprobación en escenarios adversarios.
Hoja de ruta
Firma de producción Ed25519/JWS y rotación de claves
Identidad de agente atestiguada por OAuth e identidad de carga de trabajo
Ceremonia de aprobación vinculada al hash de intención
Tokens de autorización de un solo uso para proveedores de pago
Libro mayor SQL duradero con garantías de concurrencia
Trazas de decisión OpenTelemetry y exportación de auditoría firmada
Corpus de evaluación adversaria para ataques de pago de agentes
Adaptador de sandbox Stripe/Adyen detrás de una interfaz de ejecución
Política decimal y de tipo de cambio multidivisa
Límite de uso responsable
Este repositorio es un MVP de investigación en seguridad y FinTech que utiliza identidades sintéticas y sin fondos reales. HMAC se utiliza únicamente para mantener la demostración libre de dependencias. El despliegue en producción requiere criptografía asimétrica revisada, identidad reforzada, almacenamiento transaccional duradero, autenticación y autorización, análisis regulatorio, controles de privacidad, pruebas de seguridad independientes y certificación del proveedor de pagos.
Contexto estratégico
Los sistemas de pago agénticos necesitan autorización basada en mandatos, separación entre decisión y ejecución, identidad del agente, controles programables, pistas de auditoría y supervisión humana por niveles. MandatePay convierte esos requisitos de diseño en un prototipo de código abierto inspeccionable.
Licencia
Apache License 2.0.
Available Tools
1 toolrequest_payment_authorizationA
Evaluate a proposed agent payment against a signed spending mandate. Does not move funds.
| Name | Required | Description | Default |
|---|---|---|---|
| amount | Yes | ||
| purpose | Yes | ||
| currency | Yes | ||
| merchant_id | Yes | ||
| merchant_country | Yes |
TDQS
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the burden of behavioral disclosure. The explicit statement 'Does not move funds' is a critical side-effect disclosure that prevents accidental fund transfers. However, it does not describe the output nature (e.g., approval/denial) or any other side effects, leaving some behavioral gaps. The core non-effect is well covered, so a 4 is appropriate.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is two concise sentences, leading with the core purpose and then a critical safety note. There is no wasted wording, and the most important information (the scrutiny and non-movement) is front-loaded. This is an exemplary level of conciseness.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
The description is too brief for a tool with five required parameters and no output schema. It omits the return value (e.g., whether the authorization is granted or denied), any prerequisites like having the mandate document, and does not clarify the 'purpose' field's semantics. An agent lacks critical information to interpret the tool's result and handle errors effectively, making it incomplete.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The schema description coverage is 0%, so the description must compensate by explaining the parameters. It does not describe any of the five fields (amount, currency, merchant_id, merchant_country, purpose). The context of a 'proposed agent payment' hints at their role but provides no additional meaning beyond the schema's field names, which is insufficient for a tool with no inline descriptions.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description uses a specific verb ('Evaluate') and a clear resource ('proposed agent payment against a signed spending mandate'), and explicitly states it does not move funds, which distinguishes it from payment execution tools. Even without siblings, the purpose is precise and unambiguous, leaving no doubt about what the tool does.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description clearly implies the use case: evaluating a payment proposal against a mandate before funds are moved. It does not explicitly state when not to use it or mention alternatives, but the context is strong enough that an agent understands when to invoke it. Since no siblings are present, the lack of exclusions is acceptable.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
1 tool update
v0.1.0- First observed
request_payment_authorization
TDQS
Scored across 1 tool
With only one tool, there is no possibility of confusion or overlap between tools. The tool's purpose is clearly defined and singular.
A single tool follows a consistent verb_noun pattern ('request_payment_authorization'), which is clear and descriptive. No inconsistencies exist by definition.
The server has only one tool, which feels thin for a system named 'MandatePay'—implying broader mandate management. While the tool itself is well-scoped, the overall surface is borderline minimal.
The tool covers only payment authorization, but the domain suggests the need for related operations like mandate creation, listing, updating, or checking status. Without these, agents face significant dead ends when managing mandates.
Maintenance
Related MCP Connectors
Allow, ask, or veto — a policy decision point your agent consults before signing any payment
Pre-spend firewall for AI agents. Approves, blocks, flags transactions against policy rules.
Agent payments, API key vaulting, and governed mandates. Agents spend within user-defined limits.
Check an agent action against its operator's mandate before acting: permit, deny or escalate.
Related MCP Servers
- AlicenseNot gradedqualityAmaintenanceDeterministic, auditable payment policy enforcement for AI agents. It provides pre-action authorization with scopes, budgets, allowlists, and signed mandates via an MCP server.MIT
- AlicenseNot gradedqualityCmaintenanceEnables AI agents to verify proposed payments against their assigned task, budgets, permitted categories, and counterparties, returning an ALLOW or DENY decision with a tamper-evident audit trail.MIT
- AlicenseNot gradedqualityBmaintenanceEnables AI agents to use payment APIs safely by gating every tool call through deterministic mandates, value checks, and an exactly-once ledger, so they cannot pay twice or exceed authorisation.81 npmMIT
- AlicenseNot gradedqualityFmaintenanceEnables MCP clients to ask permission before an agent acts, returning allow, deny, or ask_human with receipts, and to monitor agent spend, limits, and recent decisions.0Functional Source , Version 1.1, ALv2 Future