glm-image
Warn
Audited by Socket on Sep 22, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS: the core Pixazo API use is mostly coherent and appears to target first-party Pixazo endpoints, but the skill is broader than necessary in two ways: it tells the agent to read/store a raw API key from disk and it advertises wildcard installation of all Pixazo skills via `npx skills add ... --skill '*'`. The data flow is not outright deceptive, but the transitive-skill install and raw credential persistence make the skill medium risk overall.
Confidence: 88%Severity: 55%
Audit Metadata