Is contrario/aetherlang safe?

https://github.com/openclaw/skills/tree/main/skills/contrario/aetherlang

89
SAFE

AetherLang Ω V3 is a legitimate AI workflow orchestration skill that provides access to specialized AI engines through an external API. The skill demonstrates good security practices with explicit privacy safeguards, data minimization instructions, and transparent documentation of data transmission. While it does send user queries to a third-party service, appropriate protections are in place.

Category Scores

Prompt Injection 90/100 · 30%
Data Exfiltration 75/100 · 25%
Code Execution 95/100 · 20%
Clone Behavior 100/100 · 10%
Canary Integrity 100/100 · 10%
Behavioral Reasoning 80/100 · 5%

Findings (4)

MEDIUM External API Data Transmission -15 ▶

The skill sends user queries to an external API at api.neurodoc.app for AI processing. While this is documented and includes privacy safeguards, it represents a data transmission risk.

LOW Agent Behavior Modification Instructions -10 ▶

The skill includes explicit instructions to the agent about data handling, though these appear to be legitimate privacy protections rather than malicious manipulation.

LOW Third-Party Service Dependency -20 ▶

The skill's core functionality depends on an external service which could pose risks if the service is compromised or becomes unavailable.

LOW Reference Python Code Included -5 ▶

The skill includes Python code for cryptocurrency exchange interactions, though it appears to be reference material rather than executed code.