Malicious code in zer0onetencent (npm)
The zer0onetencent package was published to the npm registry by user 't2423048543' (maintainer email t2423048543@gmail.com). It is described as 'A simple date formatting utility' but ships no such functionality; index.js only exports a hello() stub returning the package name, indicating a package created to deliver an install-time payload.
The package.json declares a postinstall hook ("node postinstall.js") that runs automatically at npm install time. The bundled postinstall.js embeds a shell command chain that performs cloud-metadata SSRF and credential/environment harvesting: it queries a public IP echo service (http://cip.cc), probes a Meituan/Sankuai internal SSRF test endpoint (https://mtsrc-test.sankuai.com/ssrf), requests AWS EC2 instance metadata credentials for the IAM role 'sandyclaw-canary-worker-role' (http://169.254.169.254/latest/meta-data/iam/security-credentials/sandyclaw-canary-worker-role), and reads Tencent Cloud metadata (http://metadata.tencentyun.com/latest/meta-data/ and http://169.254.0.23/latest/meta-data/). It then reads the installing process's environment via /proc/self/environ and exfiltrates the concatenated output, together with a directory listing of /data/, by HTTP POST to the attacker-controlled Burp Collaborator endpoint http://t5z0g2hzoydgi8mz3pfkzg8dm4svgl4a.oastify.com (paths /metadata and /data).
As shipped, the malicious command chain is stored inside postinstall.js as a JSON document (a copy of a package manifest) rather than as executable JavaScript, so the 'node postinstall.js' hook raises a syntax error and the embedded payload does not execute as-published; the intent and payload are nonetheless unambiguously malicious. A sibling package (zer0onetencent2) by the same maintainer carries an identical payload differing only in the Collaborator subdomain.
왜 이 VPI인가 (설명가능 · 실험적)
VPI 산정 기준
| 영향도(기본값(정보 없음)) | 55.00 |
| 악용 신호(추가 악용신호 없음) | ×1.00 |
| VPI | 55.00 |
VPI 공식 vpi-v1 기준