deer-sense
Warn
Audited by Socket on Mar 18, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The stated purpose is coherent for an accessibility-audit skill, and data flows stay mostly local with no credential harvesting. However, the skill instructs the agent to execute a local `tools/glimpse` CLI whose provenance cannot be independently verified from the provided evidence, and its `npx axe` guidance is ambiguous/outdated versus the current official package. Main concern is supply-chain/install trust rather than confirmed malicious intent.
Confidence: 85%Severity: 74%
Audit Metadata