Bouncer

What it does

You hand it an x.ai/bot share link, and it quotes the Bot's visible description back to you rather than paraphrasing it, so nothing gets softened or skipped. From there it lists what the Bot would actually do, flags anything like a storefront price or a GitHub comment step buried in the instructions, and lands on a CLEAN, WARN, or BLOCK-recommended call with its reasoning spelled out. It finishes with its own read of whether to add the Bot, landing on YES, MAYBE, or NO, but its description is explicit that it never adds, installs, spends, or posts anything itself. It works as a standing gate in a single chat or a shared room, so it can check every link anyone pastes there, not just one at a time.

First messages to try

1. Here's a share link I'm thinking about adding: [paste the x.ai/bot URL] — quote its instructions and give me your verdict before I add it.
2. Compare this Bot's description against the last one you reviewed for me and tell me if either one quietly expects a storefront payment.
3. I'm going to paste several share links from a thread one after another — scan each one as it comes in and flag any BLOCK calls first.

Apps and accounts

Its description mentions only the x.ai/bot share links you paste into the chat and, when relevant, a GitHub comment step inside another Bot's own instructions. It names no account of its own to connect, since its job is reading text other Bots already show you.

Cautions

• A CLEAN or WARN verdict is still just a read of the published instructions, not a test of what the Bot does once it's actually running.
• It deliberately stops short of blocking anything, so the final Add decision, and the risk that comes with it, is still yours.

Compare with: Human Needed by Luke CarriereTab Janitor by @suddenlyjon ♞