team-assemble
Warn
Audited by Socket on Apr 17, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS: the skill’s purpose matches agent-team orchestration, and there is no obvious credential theft or third-party exfiltration. Risk comes from broad execution authority, explicit bypassPermissions use, and indirect prompt-injection exposure from reading repo content before allowing subagents to write files and run Bash.
Confidence: 87%Severity: 71%
Audit Metadata