substrate
Warn
Audited by Socket on Feb 28, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
The skill appears coherently aligned with its stated purpose of intra-agent mail management and plan review. Data flows are largely local (inboxes, identity state, plan files) with a local Web UI, which is typical for a agent coordination feature. The only notable behavioral concern is the Stop hook creating a persistent background loop to keep the agent alive, which could mask hidden activity if misused; however, within the documented purpose it serves a planed persistent-work pattern. Overall, the footprint is benign and proportionate to its stated mail/identity/plan-review role, with no evident credential harvesting, external exfiltration, or third-party dependency loading.
Confidence: 75%Severity: 75%
Audit Metadata