panel

Warn

Audited by Socket on Aug 16, 2026

1 alert found:

Security
SecurityMEDIUM
SKILL.md

SUSPICIOUS. The stated purpose matches multi-CLI orchestration, but the skill grants four external agent CLIs broad autonomous execution rights and sends them user prompts plus repository content. The main risk is not deception about purpose; it is disproportionate execution authority, external data exposure, and prompt-injection potential from running dangerous headless agents against local code.

Confidence: 90%Severity: 84%
Audit Metadata
Analyzed At
Aug 16, 2026, 04:51 AM
Package URL
pkg:socket/skills-sh/leek%2Fagent-skills%2Fpanel%2F@a6ecb16ef70627b15e65c81adebc34de5f3ac88d7f52aadffa14dcd0e7cf9a39
Security Audit — socket — panel