AgentTrust Collector
Crawler Directory evidence: Verified

AgentTrust Collector

Discovers publicly advertised AI agents and agent metadata.

AgentTrust-Collector
Operator agenttru.st
Risk Neutral

Overview

AgentTrust Collector is a specialized crawler operated by agenttru.st that discovers publicly advertised AI agents. It begins with hostnames found in public Certificate Transparency logs and requests robots.txt along with a fixed set of well-known AI agent and MCP metadata documents.

According to the operator, the crawler does not scan ports, guess arbitrary paths, follow ordinary website links, execute JavaScript, submit forms, attempt authentication, or perform vulnerability probing. A confirmed agent may be rechecked every seven days, while hosts that do not publish supported discovery documents are revisited less frequently.

AgentTrust Collector honors the agenttrust-collector robots.txt token. The operator also publishes current crawler addresses and recommends forward-confirmed reverse DNS under agenttru.st as the primary verification method.

Identity

User-Agent Pattern
AgentTrust-Collector
Aliases
AgentTrust-Collector, agenttrust-collector
HTTP Agent Examples
AgentTrust-Collector/1.0 (+https://agenttru.st/crawler; purpose=AI Agent Discovery; contact=admin@agenttru.st)
Robots Token
agenttrust-collector
Identity Type
Officially documented
Evidence Method
First-party crawler documentation and forward-confirmed reverse DNS under agenttru.st.

Classification

Type
Crawler
Kind
AI agent discovery crawler
Family
AgentTrust
Operator
agenttru.st
Company
agenttru.st
Purpose
AI agent discovery

Behavior and handling

Common Use
Discovering publicly advertised AI agents, agent cards, and MCP server metadata.
Detection Notes
Verify the requesting IP with forward-confirmed reverse DNS. The reverse hostname must end in agenttru.st, and resolving that hostname must return the original IP address. Published addresses currently include 51.83.6.2, 74.208.40.151, 54.39.22.255, and 51.38.70.50, but the operator states that the fleet may change.
Respects robots.txt
Yes
Spoofing Risk
The User-Agent can be copied by other clients. Verify legitimate requests using forward-confirmed reverse DNS under agenttru.st.
Risk
Neutral
Recommended Handling
Allow

Rules and controls

Robots.txt Snippet
User-agent: agenttrust-collector Disallow: /
Apache Rule
RewriteEngine On RewriteCond %{HTTP_USER_AGENT} AgentTrust-Collector [NC] RewriteRule .* - [F,L]
Nginx Rule
if ($http_user_agent ~* "AgentTrust-Collector") { return 403; }

Related Bots

Crawler Verified

ltx71

ltx71

ltx71
Crawler Unverified

JenkersBot

Jenkers

JenkersBot
Crawler Verified

BBBike

BBBike

BBBike