agent-reach
Warn
Audited by Socket on Jul 18, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS: the stated purpose matches internet research, but the footprint is broad and trust-heavy. Main concerns are mutable same-repo install/update paths, transitive trust in multiple external CLIs, cookie/session forwarding for login-backed platforms, and routing web fetches through `r.jina.ai` rather than direct sources.
Confidence: 86%Severity: 78%
Audit Metadata