iblai-profile
Warn
Audited by Socket on Apr 29, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS: mostly coherent developer tooling for ibl.ai profile features, with same-org installs and official-looking data flows, but it expands trust by invoking external CLI tooling and other skills, handles raw tokens from localStorage, and asks the agent to perform side-effectful local actions automatically. This is not clearly malicious, but its execution footprint is broader than a simple UI skill.
Confidence: 81%Severity: 52%
Audit Metadata