The viral 'last30days' research skill pulls Reddit, X, and web data: why it is not safe to use

What happened

A promoted community skill called `last30days` is being listed for Claude Code installation with a one line command. Its description says it pulls the last 30 days of conversation from Reddit, X, and the web to answer what people are saying right now.

The listing also says it works in different modes depending on which API keys you have. An API key is a secret code that lets one service use another service on your behalf. In this case, the listing mentions OpenAI for Reddit and xAI for X. That means a search or research task may be routed through more than one outside platform, not just Claude or Codex.

At the same time, the broader skills ecosystem has been warning users that not every listed skill is guaranteed safe.

What it means for you

For a normal user, the main issue is privacy. If you use a research skill like this, your prompts, topics, and lookups may be sent to third party services beyond the AI tool you started with. That can matter if you are researching work projects, personal matters, or anything sensitive.

What to do instead

Use community skills carefully. Read the listing, check which outside services it can call, and avoid entering sensitive prompts unless you are comfortable with that data leaving your main AI tool.

If you need research help, prefer reviewed tools with clear privacy notes and limited data sharing. AgentPod lists only reviewed, tested skills without overselling what they can do.

Sources:

• https://crossaitools.com/skills/jdrhyne/agent-skills/last30days
• https://www.skills.sh/docs