# GlintbaseResearch bot

> What the GlintbaseResearch user agent requests, what it never does, and how to opt out.

HTML version: https://glintbase.dev/research/bot

If you saw requests with this user agent, they came from Glintbase's research harness. We run small, pre-registered measurement studies of how public AI-agent infrastructure behaves for software agents.

```
GlintbaseResearch/1.0 (+https://glintbase.dev/research/bot; research@glintbase.dev)
```

## What it requests

### [Remote MCP Server Reliability](https://github.com/Glintbase/research/tree/main/benchmarks/production-mcp-reliability-crisis)

Protocol metadata from MCP endpoints listed in the official MCP Registry: server/discover or initialize, tools/list, one call to a tool that does not exist (to check error handling), and OAuth protected-resource metadata if the endpoint requires auth.

Volume: at most 8 requests per endpoint per round, one endpoint per host at a time, two rounds.

### [The Silent Wall](https://github.com/Glintbase/research/tree/main/benchmarks/the-silent-wall)

robots.txt and sitemap from 150 popular websites, then the homepage and up to 3 public pages listed in the sitemap, fetched three ways for comparison: as GlintbaseResearch (which obeys robots.txt for its own name), as a plain HTTP client with its library's default user agent, and as a headless Chromium browser with its default (HeadlessChrome) user agent. Page bodies are not stored.

Volume: about 12 page requests per site per round, at most 1 per second, with a 30-second pause between client types; two rounds.

## What it never does

- Call your real tools, or send data intended to change state.
- Send security payloads, attempt authentication, or try to bypass challenges, CAPTCHAs or rate limits.
- Pretend to be another company's crawler. Our comparison clients use their own default identifiers (a library user agent, or HeadlessChrome) and never a human browser's.
- Log in, fill in forms, or fetch pages that are not public.

## Opting out

Email research@glintbase.dev with the host name(s). We add them to the study's opt-out list and stop within 24 hours. You can also block the `GlintbaseResearch` user agent in robots.txt.

## What we publish

Aggregate statistics plus per-endpoint derived fields, such as "handshake succeeded" or "tool count". We never publish raw responses, and we don't publish lists naming the worst servers.