Security Scan Report: disconfirmai.com

Redirected to:
https://disconfirmai.com/zh/reading/reinforcement-high-agent-benchmark...
Site favicon
Submitted: Completed: publiccompleted
This website contacted 1 IP in 1 country across 1 domain to perform 28 HTTP transactions. The main domain is disconfirmai.com and was registered 2 months ago.

Submitted URL: https://disconfirmai.com/zh/reading/2026-10-20-18

Effective URL:
https://disconfirmai.com/zh/reading/reinforcement-high-agent-benchmark...
Redirected

AI Security Verdict

Low Risk

Confidence: 72%

2
Risk Score

Self-branded tech-news article about Meta's AI models; no forms, impersonation, or threat-intel hits. Only a 50-day-old unranked domain — low risk.

Risk Factors (1)
Domain is only 50 days old and has no established reputation (weak prior only)
Safety Factors (4)
Article/news page type (og:type=article) with byline-style attribution ('DisconfirmAI 编辑部') and substantive editorial content
No forms, no password or payment fields, no cross-origin credential handling
Self-branding consistent with the domain; no impersonation of another entity
Clean threat intelligence, JavaScript, and network IDS results
Domain age information unavailable

Details

Page Title

强化学习训练绑定反馈环境,Agent高分不等于陌生任务能力 | DisconfirmAI

Scan Type

public

Domain Name Analysis

The domain name 'disconfirmai.com' uses the commercial generic top-level domain (.com) while skipping any subdomain. Its registrable label 'disconfirmai' stretches across 12 characters split between five vowels and 7 consonants. It segments into five words: d, is, confirm, a, i. Median word length comes out to 1 character. No strong language cues emerged from the frequency lists.

Screenshot

@ FP EEE MR ORÆSK 3 AHKAR Disconfirmal RER RLE SNAHER DI, Agente a# STRESS BED MetaffJMuse Wondertr4i fa ktilk HAR, AMEN FER, BUGS DH AAR: FG STÅAAgent EEE SNOR IN ANT HT HE LT MAR RUE, Mi 202...

Page Load Overview

3.08s
Total Load Time
263 KB
Total Size

Language Analysis

Primary Language

🇨🇳Chinese
Code: zh
Confidence:60%
Script:Han
Direction:ltr

Detection Details

HTML Lang Attribute:zh-CN
Text Length:1,811 chars
Detector Agreement:50%

Website Classification

Primary Category

technology software59% confidence
Type: dynamic
Method: ml+structural

All Detected Categories

technology software
59%
documentation technical
45%
corporate
35%
news/blog
30%

Detected Features

OG: article
Schema.org

Domain & IP Information

RequestsIP AddressLocationAS Autonomous System
2847.237.135.51Singapore
AS45102Alibaba (US) Technology Co., Ltd.
281--

Detected Technologies2

Nginxv1.20.1
40%

Content Similarity HashesFor malware variant detection

TLSH (Trend Micro Locality Sensitive Hash)

Security-focused

Specialized for malware detection and similarity analysis

T17F43D837A008DC6FE86B6F9C117F6C3E2089867B6694C9A4D3CDCD5509458BAA2F3DC0

ssdeep (Context Triggered Piecewise Hashing)

Context-aware

Detects similar content even with modifications

768:I/FpuxoI4abcjVz/YwxaUcqo7TCAC0SSvHhHwZfhQinNI+IC+aop6+98ktkcg:R8AEcmHYM/

sdhash (Similarity Digest Hashing)

High-precision

High-precision similarity detection for forensic analysis

sdhash:3:57371:AQAQQEhcUVRDjGDARAIkDNIUKSVTSmVqIEwCAzDRYAtA5AO60AWJCgSgCDhEiZ7RnujEJgDIIAOAIKTAuJGohhhBgIkgYRik

These hashes enable detection of similar websites and malware variants by comparing content similarity even when exact matches aren't found.

Image Hashes

Perceptual Hashes

Average Hash:efc3efe7c3c3c3c3
Perceptual Hash:b1993066ccce8ece
Difference Hash:9e9e9e869e9e8e96
Wavelet Hash:7fc2c2c2c3c3c3c3
Color Hash:N/A

Other Hashes

Crop Resistant:9e9e9e869e9e8e96

Scan History

Scan history not available

Unable to load historical scan data