autoresearch
Warn
Audited by Socket on Mar 29, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS: The skill’s core purpose matches its capabilities, but it normalizes high-autonomy operation with --allow-all, unattended overnight experimentation, and automatic repository changes. The Copilot CLI provenance appears official, so supply-chain concern is limited; the main risk is autonomous modification of prompt/code assets with broad local execution and write access.
Confidence: 89%Severity: 78%
Audit Metadata