Trends MCP: why it is not safe to use
Trends MCP passed our review; we found a disclosed hosted read-only service with no hidden instructions, hardcoded secrets, or risky commands.
What we found
Our review did not identify a security or privacy issue that would block Trends MCP. Based on the inspected repository materials, the skill presents itself as a hosted MCP and REST service at `trendsmcp.ai` and `api.trendsmcp.ai`, so user queries are sent to the service the user intentionally connects. We did not see hidden prompt instructions, hardcoded credentials, obfuscated payloads, undisclosed remote code fetches, or install steps that attempt privilege escalation.
The documented behavior appears limited to read-only trend lookup. The examples show query-style calls, with no delete, overwrite, publish, or other irreversible actions. We also did not find evidence in the inspected materials that it accesses unrelated environment variables, local keychains, browser data, or SSH keys beyond the API key or OAuth session used for the Trends MCP connection.
What to do instead
If you choose to use this skill, treat it as a third-party hosted service and share only the query data you intend to send to `trendsmcp.ai`. Use a scoped API key or OAuth session, follow the repo guidance not to commit secrets, and review the provider's own privacy and retention terms before connecting it in production.
Source: https://github.com/trendsmcp-ai/Trends-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.