ic-eo-interpretable-code-based-assistant
Warn
Audited by Socket on Jun 20, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS: the skill’s EO purpose is coherent, but it gives an AI agent a code-generation-and-execution loop over untrusted user input with networked geospatial data access. There is no clear malicious exfiltration path, yet the combination of executable code synthesis, implicit credential scope, and unspecified `iceo_api` provenance creates meaningful security risk.
Confidence: 84%Severity: 64%
Audit Metadata