experimentation-engineer
Warn
Audited by Socket on Aug 27, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS. The stated purpose is benign and the skill text itself contains no direct malware, exfiltration endpoint, or credential-harvesting instruction. However, it mandates several external tools/services and a command proxy (rtk) that are not scoped or pinned in the skill, creating medium supply-chain and execution-trust risk disproportionate to a planning-oriented experimentation skill. Data flows are not overtly malicious, but the combination of untrusted web input, command execution, and persistent memory writes increases exposure.
Confidence: 88%Severity: 52%
Audit Metadata