convez.ai
MCP ServerSegurançaAvançado
dinpd

AgentPass

dinpd/AgentPass

Autorização just-in-time para ações de agentes IA em sistemas críticos

Ver no GitHub
InicialPluginScore · de 0 a 100

MCP server que implementa autorização em tempo de execução para chamadas de ferramentas de agentes de IA. Funciona como uma porta lógica que valida se um agente pode executar uma ação específica em um recurso, permitindo, negando ou desafiando a requisição antes que o agente toque em sistemas de produção, dados sensíveis ou transações financeiras.

Instalação

$npx convez-ai dinpd/AgentPass

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
7
Forks
0
Acompanhando
0
Saúde
42%
Último commit
há 6 dias
Versão
v0.60.0
Issues abertas
37
Contribuidores
1

Casos de uso

  • Controlar execução de ações de agentes em sistemas SaaS e APIs
  • Implementar políticas de aprovação e desafio para operações sensíveis
  • Auditar e registrar todas as tentativas de execução de ferramentas por agentes
  • Proteger dados de clientes e operações financeiras contra agentes descontrolados

Veredito

Solução madura e bem-arquitetada para um problema real em produção: agentes de IA precisam de controle de autorização granular antes de tocar em recursos críticos. AgentPass oferece múltiplos pontos de entrada (guard local, gateway MCP, policies enterprise) e prioriza auditoria, sem substituir IAM ou ferramentas de segurança existentes—complementa-as.

README

A tradução para o português deste README ainda está na fila.

AgentPass

Runtime authorization for AI agent tool calls.

AgentPass is a runtime action gate that sits before agents call tools, send messages, move money, update records, deploy code, export data, or touch sensitive systems.

Agent proposes tool call -> AgentPass checks policy -> allow / deny / challenge

AgentPass helps builders ship bounded AI automations and agents without giving models unchecked execution authority over production systems, customer data, payments, external sends, secrets, and administrative tools.

Explainer Video

https://github.com/user-attachments/assets/4d5757f7-ffa0-4c0f-a5bc-1bde4336703a

AgentPass answers one runtime question:

Should this agent perform this capability or tool action on this resource, for this user, job, customer, approval, and time window?

AgentPass is one control model with different entry points depending on where you sit in the agent stack:

  • Agent developers start with the local TypeScript guard to wrap tool calls in an existing agent loop.
  • Enterprise AI and platform teams use manifests, approvals, gateways, policy checks, and audit to govern agents across systems.
  • MCP gateway builders enforce AgentPass before forwarding tools/call.
  • API and SaaS providers publish provider contracts and verify scoped receipts before executing agent-originated actions.
  • Security and risk teams review action policy, PII/data-flow controls, decision events, JIT grants, and kill-switch behavior.

The first package is the local TypeScript guard:

The guard package is published on npm as @dinpd/ai-agent-guard. The repo also includes local demos, starter policies, and the MCP tool-call adapter so developers can validate the policy shape inside an existing agent loop before adding broader gateway or enterprise enforcement.

OAuth can prove access to a server. MCP tool schemas describe inputs. Agent frameworks can decide which tools are visible to a model. AgentPass focuses on the runtime decision immediately before a tool executes.

The core idea is simple:

Every production agent should have a runtime action contract that says what it can request, what must be challenged, where sensitive data can flow, which retries are safe, and how execution should be stopped.

AgentPass does not replace IAM, OAuth, MCP gateways, OPA, Cedar, or enterprise security tools. It gives agent runtimes a small policy checkpoint for tool calls, approvals, data movement, circuit breakers, and audit events.

Core Concepts

AgentPass separates three concepts that are often blurred in agent systems:

Skill = workflow package
Tool = executable operation
Flow = data movement boundary
  • Tool: an operation the agent or runtime can call, such as provider.crm.search_customer, stripe.create_refund, or email.send_external. Tool policy answers what operation is being attempted, what access level it has, which resource it affects, and whether approval is required.
  • Skill: a reusable workflow package that may call one or more tools, such as support-refund-workflow. Skill policy becomes relevant when reusable workflows need explicit downstream tool limits.
  • Flow: a source-to-destination data boundary, such as provider_crm -> agent_context or customer_records -> external_email. Flow policy answers where data may move and which destinations are blocked.

The local guard implements tool and flow checks first. Larger deployments can apply the same policy model through manifests, signed receipts, hosted gateways, and provider-side verification.

Entry Points By Audience

  • Platform engineering and SRE teams letting agents inspect systems while gating production deploys, rollbacks, Terraform applies, Kubernetes changes, incident remediation, secrets, and IAM changes.
  • Enterprise AI platform teams reviewing which tools agents may use, under what conditions.
  • Security teams needing approval, audit, and kill-switch evidence for high-risk agent actions.
  • MCP gateway builders enforcing policy before forwarding tools/call.
  • API and SaaS providers turning APIs into MCP tools without giving agents broad authority.
  • API platform and monetization teams preserving entitlements, quotas, metering, and billing controls as APIs become agent-callable tools.
  • Skill authors and platform teams packaging reusable workflows with explicit AgentPass guardrails.

For gateway deployments, AgentPass can run at an enterprise-controlled boundary as the authorization decision service, not necessarily as the network gateway or MCP proxy:

Enterprise Agent -> Enterprise Gateway or App Runtime -> AgentPass Check -> Internal, SaaS, or MCP Tool

If you are an agent developer:

If you are evaluating enterprise governance, gateway enforcement, provider contracts, receipts, or standards alignment:


Quick Start

Try the first developer wedge: add a circuit breaker and approval gate before your agent executes tools.

Install the package:

npm install @dinpd/ai-agent-guard

Run the local demos:

git clone https://github.com/dinpd/AgentPass.git
cd AgentPass/packages/guard
npm install
npm run demo:quickstart
npm run demo:mcp
npm test
npm run demo:circuit
npm run demo:gate
npm run demo:pii

Use it in an agent loop:

import { createToolGate } from "@dinpd/ai-agent-guard";

const gate = createToolGate({ policy });

const execution = await gate.run(
  {
    agentId: "support-agent",
    jobId: "case-1042",
    tool: "stripe.refund",
    action: "pay",
    resource: "payment/pi_123",
    amountUsd: 49,
    idempotencyKey: "refund-case-1042-pi_123"
  },
  () => stripe.refunds.create({ payment_intent: "pi_123", amount: 4900 })
);

if (!execution.executed) {
  return execution.decision;
}

The guard currently demonstrates:

  • Circuit breakers for runaway tool calls, repeated calls, token spend, cost, and runtime.
  • Action controls for approvals, amount caps, and single-use idempotency.
  • PII/data-flow controls for approved destinations, blocked fields, record counts, and model-provider prompts.
  • Audit events for every allow, deny, or challenge decision.

Starter policies:

Enterprise manifest and provider-contract tooling is available when you need a reviewable governance layer beyond the local guard package:

git clone https://github.com/dinpd/AgentPass.git
cd AgentPass
python -m pip install -e ".[dev]"
agentpass validate examples/provider-mcp-support-agent.yaml
agentpass risk-score examples/provider-mcp-support-agent.yaml
agentpass generate-policy examples/provider-mcp-support-agent.yaml --target opa

AgentPass installs agentpass as the primary CLI and keeps agentid as a compatibility command alias. The Python package, schema filenames, environment variables, and receipt field names still use agentid for compatibility.

Feedback Wanted

This package is intentionally local-first. Before adding hosted gateways or account infrastructure, the next useful feedback is:

  • Which failure mode matters most in your agents: token spend, tool loops, duplicate side effects, PII exfiltration, shell/browser actions, or something else?
  • Does allow / deny / challenge_required fit how your agent runtime handles tool calls?
  • What fields are missing from the guard check: tenant, customer, workspace, channel, model, tool params, MCP server, approval source, or policy version?
  • Which adapter should come first: OpenClaw, OpenAI Agents SDK, LangGraph, CrewAI, AutoGen, MCP gateway, or plain Express/FastAPI middleware?
  • Would you rather run this as an in-process package, a local sidecar, or a hosted policy service?

Broader Governance Scope

The same action-gate model can move from an in-process guard to shared boundaries such as app runtimes, MCP gateways, provider MCP servers, and security-controlled policy services.

For enterprise governance, gateway, provider-contract, receipt, and standards documentation, see Enterprise Governance.

For the runtime guard roadmap, see Action Gate Roadmap.


License

Apache-2.0

Ver o original no GitHub

Bom para

Relacionados