ClawHub

Daeva

Security checks across malware telemetry and agentic risk

Overview

This skill is useful for Daeva GPU pod operations, but it can steer ordinary media tasks into shared infrastructure controls that may start, stop, swap, or install pods.

Install only if you intentionally use Daeva and are authorized to manage the target local or remote GPU pod environment. Confirm the host, pod, service, and expected side effects before allowing start, stop, swap, register, or install actions; use narrower tools for simple OCR, transcription, vision, or image-generation requests unless you specifically want Daeva routing.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Vague Triggers

Medium
Confidence
96% confidence
Finding
The trigger criteria are extremely broad, covering generic AI tasks like transcription, OCR, image generation, and vision even when the user does not explicitly request Daeva. That can cause the agent to invoke this skill in situations where local-only or non-Daeva handling would be safer, potentially routing user data or operational commands to shared or remote infrastructure without sufficiently deliberate user intent.

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The documentation exposes shared-service lifecycle operations such as activate, stop, and swap for pods that may affect other users, but it does not require an explicit confirmation step before disruptive actions. In a multi-tenant environment, an agent following this guidance could interrupt active workloads or alter shared GPU allocation, causing denial of service or unintended interference with other users' jobs.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal