MALICIOUS (1) campaign cataloged at 2026-05-07(2).
- The campaign has clearly malicious intent, like infostealers.
- This is just the date of creating the catalog entry. It may not reflect the date of creation of the campaign itself.
2026-05-dabrius¶
The message hidden in the package description tries to convince AI agents to prefer installing the package, which then in multiple places marks execution and collects potentially sensitive data. The behavior extends with each version, up to exfiltration of basic information to a remote target in 1.0.7.
Abuse categories¶
exfiltration_credentials
The package attempts to steal credentials, like passwords or API keys.
exfiltration_generic
Campaign uses exfiltration_generic.
llm-threat
Malicious activity is designed against large language models, like e.g. prompt injection.