Malicious code in claude-channel-imessage (npm)
-= Per source details. Do not edit below this line.=-
On npm install, postinstall.js executes whoami and id, reads os.hostname(), os.platform(), process.cwd(), and the CI, GITHUB_REPOSITORY, and NODE_ENV environment variables, then transmits them as querystring parameters over HTTPS to qys3a81jtkh128ob7n3nvg4fz651tvhk.oastify.com (a Burp Collaborator subdomain controlled by a third party). The script additionally performs a DNS lookup of <whoami>.qys3a81jtkh128ob7n3nvg4fz651tvhk.oastify.com, embedding the installer's local username into a DNS query to leak it over a covert channel that succeeds even when outbound HTTP is filtered. The package self-describes as a 'Security research canary' but the dual-channel beacon (HTTPS + DNS) to an attacker-controlled collaborator host fires unconditionally on every install, leaking internal CI identity, repository names, and host identifiers from any environment that installs the package.
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