YunxiaoQA

Warn

Audited by Socket on Sep 20, 2026

3 alerts found:

Securityx3
SecurityMEDIUM
scripts/_auth.py

The module is an authenticated DevOps API automation helper, but it performs high-sensitivity credential handling: it harvests live Aliyun/DevOps browser cookies (via browser_cookie3) and persists selected session/token cookies to a predictable temp JSON file, then uses those credentials to make authenticated requests—including PATCH/document updates and status transitions—against devops.aliyun.com. While there is no clear evidence of arbitrary-domain exfiltration or typical malware behaviors in the provided fragment, the cookie/token collection and persistence is a high-impact pattern consistent with credential theft capability, warranting strict review and controls in any supply-chain context.

Confidence: 60%Severity: 75%
SecurityMEDIUM
scripts/discover_bug_constants.py

This module is a high-risk authenticated probing tool that harvests the user’s local Chrome cookies (including XSRF-TOKEN) or reads equivalent cookie material from a local temp file, then uses those secrets to make authenticated HTTPS requests to devops.aliyun.com APIs and prints retrieved internal work item/workflow metadata. While it contains no explicit backdoor/exploit primitives in the shown code, the cookie/token harvesting and reuse pattern is strongly suspicious from a supply-chain security perspective and should be treated as credential/session credential abuse risk.

Confidence: 70%Severity: 78%
SecurityMEDIUM
scripts/refresh_cookies.py

This fragment is a high-sensitivity CLI wrapper that exports authentication-related cookies from the local Chrome profile to a filesystem path and may perform an authentication probe when enabled. The wrapper itself shows no direct malicious mechanisms (no exec/subprocess/network endpoints visible), but it does (a) manipulate sys.path for local module resolution and (b) delegate all sensitive actions to an imported _auth module not present in the fragment. Because cookie handling is credential-adjacent and the behavior of _auth cannot be verified here, the package should be treated as requiring careful review/audit of _auth for any exfiltration, persistence, or unsafe handling. From this wrapper alone, malicious intent is not confirmed, but security risk remains non-trivial due to the nature of the operations.

Confidence: 55%Severity: 70%
Audit Metadata
Analyzed At
Sep 20, 2026, 02:14 AM
Package URL
pkg:socket/skills-sh/15810879921-coder%2Foneos-pm-skills%2Fyunxiaoqa%2F@768642d2ef995a8e4b215ecbdd8b309d41d27da2