agent-discord
Warn
Audited by Socket on May 8, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The messaging capabilities fit the stated purpose, but the skill's main auth model—silent extraction of Discord user tokens from desktop/browser storage—is disproportionate and high-risk. It also enables autonomous posting and file uploads on behalf of the user. Install path looks more like ordinary npm distribution than malware, but the credential-harvesting-style auth flow and broad account access make this skill risky.
Confidence: 89%Severity: 84%
Audit Metadata