deep-research-pro
Warn
Audited by Socket on Mar 20, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The stated purpose matches web research, but the footprint includes a required unverifiable local search executable and broad ingestion of untrusted web content with write/delegation capability. No clear credential theft is shown, but install trust is weak and prompt-injection exposure is materially high for an AI agent skill.
Confidence: 89%Severity: 78%
Audit Metadata