AgentPod is building a private, secure device for your AI agent.The AgentPod device is coming.Coming soonBe first
We checked this and rejected itprivacy

bank-mcp: why it is not safe to use

We found no hidden code or write actions, but this skill stores sensitive bank API credentials and certificates locally, which warrants caution.

What we found

Our review of `bank-mcp` did not find hidden instructions, obfuscated code, install-time fetch or execute hooks, hardcoded secrets, or privileged system changes. The exposed MCP tools appear limited to read-only banking functions such as listing accounts, listing transactions, and checking balances. The documented data flow also indicates direct communication between the local MCP process, the user’s chosen bank API, and the MCP client, without a stated cloud relay or telemetry layer.

The reason for our warning is narrower: this skill stores and reads sensitive banking API credentials and certificates from the local filesystem under `~/.bank-mcp` so it can connect to a user’s bank accounts. That behavior matches the skill’s stated purpose, but it involves highly sensitive financial access material and should be clearly understood before installation.

What to do instead

Only use this skill if you are comfortable keeping banking credentials and certificates on the device where it runs. Review the local storage path and file permissions during setup, use a dedicated low-risk bank API account where possible, and avoid installing it on shared or unmanaged machines.

Want the same outcome, safely? Use our checked skill instead.

Source: https://github.com/elcukro/bank-mcp

We report what our security review found at the time we checked, with the goal of keeping people safe. Projects change; if a maintainer has since fixed this, we are glad to recheck it. Email hello@agentpod.com.

Copied to clipboard. Paste it into your AI (ChatGPT, Claude, or your agent) to add the skill.