paperclip

Warn

Audited by Socket on Jul 31, 2026

1 alert found:

Security
SecurityMEDIUM
SKILL.md

SUSPICIOUS: the skill’s purpose mostly matches its capabilities, but its trust model is weak. The main issue is a remote-installed, insufficiently verifiable Paperclip executable that then receives API credentials and mediates access to remote corpora; optional upload/share/fetch features add additional data-egress risk even though the instructions try to scope them to explicit user intent.

Confidence: 89%Severity: 84%
Audit Metadata
Analyzed At
Jul 31, 2026, 06:21 PM
Package URL
pkg:socket/skills-sh/K-Dense-AI%2Fscientific-agent-skills%2Fpaperclip%2F@8c3901e891f9aa89c26a74d4d87a83c4ab33a3c3b9693c2178c6b928cd7b5633
Security Audit — socket — paperclip