research
Warn
Audited by Socket on Mar 19, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The research capability is broadly coherent, but the skill adds an automatic Notion upload path that forwards NOTION_API_KEY to an unverified local script under ~/.claude/skills. Because the uploader's provenance and remote endpoint are not verifiable from the skill, the credential-forwarding and data-flow risks are disproportionate to a research skill.
Confidence: 84%Severity: 82%
Audit Metadata