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

Grocery CLI MCP: why it is not safe to use

We found local handling and caching of store session tokens and pasted Cookie headers, which raises privacy risk despite otherwise clear controls.

What we found

Our review did not find hidden prompts, hardcoded secrets, obfuscated code, remote code fetching, or signs of privilege escalation. The bundled skill is explicit about when it should activate, requires a priced plan, and asks for approval before cart changes. It also states that it does not place orders.

The main reason this skill did not fully pass is narrower: for some supported stores, it legitimately works with sensitive account material such as login tokens or browser Cookie headers pasted by the user, and it caches per store session data locally. The materials we reviewed also say the user logs in themselves and that the skill does not handle the password directly. Even so, locally stored session data and pasted cookies can still be sensitive because they may provide account access if exposed on the device.

What to do instead

Use this skill only on a device and account profile you trust. Prefer stores and login methods that avoid pasting browser cookies where possible. Review how local session data is stored, clear cached sessions when finished, and avoid using it on shared machines or with higher value retail accounts.

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

Source: https://github.com/jgalea/grocery-cli

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.