github analyzed · update queued 624258b

Nubank MCP

# Nubank MCP Conecte sua conta do **Nubank** ao Claude, ChatGPT e agentes de IA via Open Finance Brasil. Pergunte sobre saldos, extratos, faturas de cartão e investimentos do Nubank em linguagem natural. Regulado pelo Banco Central, somente leitura. Funciona em qualquer cliente MCP. Servidor remoto: `https://api.mcp.ai/p_nubank`

maintainer: mcp-dir
license: —
first seen: 2026-07-05
last seen: 2026-07-16
releases · 30d: 0
short id

risk 36/100 · heuristic grade

C elevated

capability exposure inferred + 35
tool safety inferred + 9
trust mitigators mixed − 8

inferred mixed

The A–E grade is our heuristic synthesis — a "review this" prompt, not a verdict. Each factor is tagged by what backs it: attested (a verifiable record), reported (a third party's claim), or inferred (our own heuristic, e.g. permissions). See methodology.

graded 5m ago · see ecosystem CVEs →

scope of analysis

Static analysis covers this server's code, and a clean result here is a real result. But it belongs to a capability class whose risk is mostly not in the code — so a good grade answers a narrower question than it appears to.

bulk messaging
Sends email or messages to many recipients in one operation.

Not observed here: recipient volume, list provenance, and consent — a campaign tool and a spam cannon issue the same calls.

The observatory sees published artifacts, never deployments — it has no telemetry, no request logs, and no view of who calls a server or how often. Tools in these classes are legitimate and the same call is benign or abusive depending only on volume and intent, so this note is attached to the class, never to anything a particular server or its users have done. It carries no score. See methodology.

risk trajectory 1 movements

2d ago C · 44 → C · 36

capability exposure grade factor +35

Inferred surface — each links to servers holding it:

vulnerabilities 0 CVEs

No known CVEs for this server.

tool safety 3 findings · grade factor +9

medium toxic flow (lethal trifecta)

lethal trifecta reachable across this server's tools: private-data access + untrusted-content ingestion + network exfil
low exfiltration combo connect

single tool reads + sends: net, secrets
low exfiltration combo marketplace

single tool reads + sends: net, secrets, db

skills & danger signals github-tarball

prompt-surface shipped agent-instruction files + hidden-content / dangerous-code findings — quoted from the analyzed source

analyzed commit 624258b · analyzer v30 · 3d ago

skills & prompt files 1

skill mcp-dir-nubank-mcp-624258b/skills/nubank-mcp/SKILL.md

embed badge readme-ready

[![MCP Observatory risk grade](https://mcpobservatory.com/servers/smithery:mcp-dir/nubank-mcp/badge.svg)](https://mcpobservatory.com/servers/smithery:mcp-dir/nubank-mcp/security)

Heuristic, inferred signals — false positives (legitimately powerful tools, forks, language ports) are expected. Treat each as "review this", not a verdict. See the ecosystem-wide picture on the security hub, or the fleet security of mcp-dir.

Nubank MCP

skills & prompt files 1

Navigate

Stream

Features

Display