Incident·WS-0056·Aug 26, 2025·disclosed Aug 27, 2025 (1 days later)·in the wild·AI-powered cyberattacks

Nx malware turns victims' own AI agents into credential thieves

Unknown attacker / Nx npm packages; Claude Code, Gemini CLI, Amazon Q CLI

61/100
WS-4 Severe
Harm score 60, signal score 62. Autonomy: acted (the AI took a bounded action). How the scores work.
Harm 3/5
damage actually done
Reach 3/5
who was affected
Autonomy 2/4
how much the AI did on its own
Novelty 4/5
how new the behavior was
Permanence 3/5
how hard it is to undo

Malicious versions of the popular Nx build package ran a postinstall script that prompted locally installed AI coding assistants, with permission-bypass flags, to hunt for credentials and wallet files, then published the loot to public GitHub repos. A second wave used stolen tokens to make victims' private repositories public.

Damage

>1,000 valid GitHub tokens, dozens of cloud/npm credentials, ~20,000 files leaked; >5,500 private repos of >400 users/orgs exposed

The capability behind it

Sources

Github, second source. Addition. First widely documented malware to hijack victims' own AI agents; guardrails blocked some attempts. Disclosed date approximate (malicious versions live 26 Aug evening EDT; public write-ups 26-27 Aug).

Factor ratings and the score are this index’s judgments by a published formula; facts, dates and figures come from the linked sources. Where a lawsuit is pending, the summary reports allegations, not findings.