ZAST AI

@zastai@infosec.exchange
4 Followers
1 Following
27 Posts
AI security researcher,
who can find zero-days with zero false positives,
as fast as other AI developers write code.
Public Relationpr@zast.ai
Websitehttps://zast.ai/

Alert: Unauthenticated Arbitrary File Upload leading to RCE.
ZAST engine has identified a critical-severity vulnerability, CVE-2026-1405 (CVSS 9.8), in the Slider Future WordPress plugin. This flaw allows for Unrestricted Arbitrary File Upload, leading to full Remote Code Execution (RCE).

Key Technical Findings:
- Vulnerability: Unauthenticated Arbitrary File Upload to RCE
- Project Popularity: 1,000+ active installations.
- Verification: 100% verified via Autonomous PoC generation.

The vulnerability stems from a lack of authentication on the /wp-json/slider-future/v1/upload-image/ endpoint and a total absence of file type or content validation before writing to disk.

We have verified that an attacker can upload a malicious PHP script and gain control of the host server in seconds.

Check detail here:https://www.cve.org/CVERecord?id=CVE-2026-1405

@wordpress@lemmy.world @WordPress @wordfence

#AppSec #ZAST #VulnerabilityResearch #WordPress #RCE

ZAST Security Advisory: Critical SSRF Resolved in ClawdBot.

While the community focused on general configuration risks, ZAST verified the actual code.

Our engine autonomously identified a high-severity SSRF vulnerability exploitable via DNS Rebinding. The flaw was a classic TOCTOU (Time-of-Check to Time-of-Use) gap, allowing attackers to bypass validation and access internal networks.

The Resolution: Our Co-founder @beach reported this to the maintainer @steipete, who acknowledged the issue and pushed a fix immediately.

The project has now implemented DNS Pinning to eliminate the vector. We are proud to be credited in the changelog for securing the ecosystem.

View the official fix: https://github.com/clawdbot/clawdbot/commit/b623557a2ec7e271bda003eb3ac33fbb2e218505

#AppSec #CyberSecurity #SSRF #Clawdbot #OpenSource #ZAST

We verified a Stored XSS (CVE-2026-0693) in the "Allow HTML in Category Descriptions" @WordPress plugin.

The Flaw: The plugin correctly restricts input but unintentionally removes global output filters (wp_kses_data) for all users. The Impact: Malicious scripts in category descriptions execute for any visitor. The Validation: Confirmed via autonomous PoC.

Security requires validating the full data lifecycle, not just lines of code.

Vulnerability details: https://www.cve.org/CVERecord?id=CVE-2026-0693
@wordfence @cve @zoomeye_team

#WordPressSecurity #AppSec #ZAST

test
😫Tired of fake vulnerabilities + slow checks?
Zast.ai fixes it—with actual working POCs!
⚡Talk is cheap, show me the POC!
Dive in: https://tinyurl.com/yjz8f5nd
#Cybersecurity #AI #POC #Vulnerability #Automation
Vulnerability Assessments Without PoC Are a Waste of Time!

Discover why vulnerability assessments without a Proof of Concept (PoC) are a waste of time. This article explores how AI and LLMs can automatically generate effective PoCs for vulnerabilities like SSRF and Path Injection, enabling teams to validate real risks quickly and eliminate false positives.

Zast.ai

[Sneak Peek] Progress Update
🎯 Zast.ai now can find 0-day in Python code with ZERO false positives - verified with a working PoC.

🐍 Python - IN BETA, full release coming soon!

✅ JavaScript - Production ready

✅ Java - Production ready

More languages support is on the way, stay tuned for the official launch!

#Cybersecurity #AI #DevSecOps #Python #ZAST

🎯Hundreds of zero-day vulnerabilities from dozens of open-source projects. By AI agent: Zast.ai.

We've just proven large-scale vulnerability discovery is not only possible, it's devastatingly effective. But the responsible disclosure story? Buckle up 👇
https://tinyurl.com/ycxa4cme

#ZeroDay #AppSec #Automation #OpenSourceSecurity

Finding Zero-Day Vulnerabilities at Scale: Our Journey with Zast.ai

Learn about our journey using Zast.ai to discover hundreds of zero-day vulnerabilities across the open-source ecosystem at scale, and the challenges we faced in responsibly disclosing them.

Zast.ai
🎯Hundreds of zero-day vulnerabilities from dozens of open-source projects. By AI agent: Zast.ai.
We've just proven large-scale vulnerability discovery is not only possible, it's devastatingly effective. But the responsible disclosure story? Buckle up 👇

https://blog.zast.ai/security/ai/open%20source/Finding-Zero-Day-Vulnerabilities-at-Scale/

🚀 Zast.ai is live!

We've developed an AI agent achieving ZERO false positives in vulnerability assessment.

Real results:
• Top contributer to the vulnerability database VulDB.com in July, 2025, with 78 submissions in 12 days.
• 300+ zero-day vulnerabilities uncovered so far
• Every vulnerability verified with working PoCs

🔗 Learn more: https://blog.zast.ai/security/ai/vulnerability%20assessment/Introduce-Zast.ai/

The future of security is here.

#Cybersecurity #AI #ZeroDay

A Shared Pursuit: Introducing Zast.ai

Introducing Zast.ai - AI agent that can identify vulnerabilities, and verify exploitability with zero false positives. Join us in making software more secure.

Zast.ai