autoresearch
Warn
Audited by Socket on Oct 9, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS: the skill's purpose matches autonomous experimentation, but its nonstop autonomy and broad permission to modify code, run shell commands, commit changes, and perform destructive cleanup create high operational risk. There is no clear credential theft or external exfiltration, so this is not malware, but it is a high-risk autonomous workflow skill.
Confidence: 90%Severity: 76%
Audit Metadata