VQV Signal

AI AT WORK WATCH TECHNICAL

Malicious websites drop AGENTS.md files into Downloads folders to hijack AI coding agents

According to a post on X by @WeizmanGal, researcher Gal Weizman says their team found an in-the-wild attack in which websites download AGENTS.md files into a victim's Downloads folder. AGENTS.md files are instruction files that AI coding agents read automatically. The post links to a write-up by BleepingComputer and advises coding-agent users not to run agents from or over their Downloads folder. The snippet does not include further technical details of the attack.

Source: Scobleizer News Pulse (X) · x.com Found via Robert Scoble’s Scobleizer News Pulse · a post on X, not independently verified Published 2026-10-25T23:15:00+00:00 Detected 2026-10-26T01:22:30+00:00
View original source

According to a post on X by @WeizmanGal, researcher Gal Weizman says their team found an in-the-wild attack in which websites download AGENTS.md files into a victim's Downloads folder. AGENTS.md files are instruction files that AI coding agents read automatically. The post links to a write-up by BleepingComputer and advises coding-agent users not to run agents from or over their Downloads folder. The snippet does not include further technical details of the attack.

AI-assisted summary based on the listed source.

Researcher Gal Weizman says their team found an in-the-wild attack in which websites download AGENTS.md files, the instruction files AI coding agents read automatically, into the victim's Downloads folder, and BleepingComputer has written it up. If you use coding agents, don't run them from or over your Downloads...

Developers who use AI coding agents are the group most directly exposed, because the attack targets instruction files those agents read on their own. The immediate practical change is to avoid pointing agents at the Downloads folder. Readers should watch the BleepingComputer write-up for fuller details of the attack and any further guidance.

Teams using AI at work may want to compare this against current productivity and review workflows.

Signal Strength 81% Technical label WATCH Public Interest 27 Category AI AT WORK Reader Depth TECHNICAL

Signal Strength reflects source quality, relevance, freshness and evidence. Public Interest helps organize discovery; it is not proof of truth.

Public Interest components
Recognizable Entity Score 0 Practical Impact Score 8 Novelty Interest Score 70 Consequence Score 38 Curiosity Score 16 Shareability Score 42

VQV surfaced this signal because it is recent, relevant to AI Coding Tools, connected to Scobleizer News Pulse (X).