agent-slack
Warn
Audited by Socket on May 15, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The skill’s Slack-management capabilities mostly match its stated purpose, and npm-based installation is not inherently abnormal. However, the authentication design is high risk: it silently extracts Slack user tokens/cookies from the desktop app or browser, stores credentials locally, and then enables an AI agent to perform broad real-world actions in the workspace. This is a coherent but overly powerful and intrusive design with substantial credential-handling and autonomy risk.
Confidence: 88%Severity: 84%
Audit Metadata