Malicious code in xinference (PyPI)
MAL-2026-3000
Published · Modified
Dependency scanning
Check whether xinference is in your codebase
Corgea flags malicious and compromised dependencies with reachability analysis, so you fix the packages that actually run in your application instead of working through the whole lockfile.
Description
__
Source: kam193 (1d006f6a08c959393160456d4ace221fd165b6d609fc8356ebfb041979aef93d)
Versions 2.6.0, 2.6.1, 2.6.2 were compromised.
Following a malicious pull request that exfiltrated sensitive data from the CI runner, three malicious PyPI releases were published. Infected releases contain code typical for TeamPCP actions that exfiltrates all kinds of sensitive data (credentials, env variables, SSH keys, cloud tokens, configuration files, shell histories, cryptowallets, data from secret managers...). Malicious action activates during importing the main package's module. TeamPCP denies their involvement.
Category: MALICIOUS - The campaign has clearly malicious intent, like infostealers.
Campaign: 2026-04-teampcp
Reasons (based on the campaign):
exfiltration-env-variables
exfiltration-ssh-keys
obfuscation
exfiltration-cloud-tokens
exfiltration-crypto
exfiltration-credentials
compromised-package
exploited-ci-vulnerability
References
- WEB https://github.com/xorbitsai/inference/issues/4828
- WEB https://github.com/xorbitsai/inference/pull/4826
- WEB https://github.com/oaoaoaoaooaoa/inference/blob/f6cf73a913443d8098d7fe2456b54cdc25eb9574/doc/source/gen_docs.py
- WEB https://bad-packages.kam193.eu/pypi/campaign/2026-04-teampcp
- WEB https://x.com/pcpcats/status/2046927940932260092
- WEB https://research.jfrog.com/post/xinference-compromise/
Ready to move
Start Securing
Free, no credit card | First findings in minutes