Security audit

Free AI Bot

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed AI provider router; it may send prompts to Cloudflare or Groq when cloud providers are used, but that behavior fits its stated purpose.

Install only if you are comfortable with an AI router that can send prompts to Cloudflare or Groq when cloud providers are selected or auto fallback reaches them. Use the ollama provider for local-only prompts, avoid sending confidential data to cloud providers unless acceptable, and keep Cloudflare/Groq API tokens scoped and out of shared logs or repositories.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (5)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 88% confidence
Finding: The skill advertises and relies on environment variables and outbound network access, but does not declare corresponding permissions. This creates a transparency and consent problem: users or hosting platforms may not realize the skill can read secrets from the environment and send data externally, which increases the risk of unintended credential exposure or unauthorized data egress.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The README promotes 'intelligent routing' and failover across local and cloud providers, but it does not clearly warn users that prompts may be sent to third-party services such as Cloudflare or Groq. This can mislead privacy-conscious users into assuming requests stay local, causing accidental disclosure of sensitive prompts, data, or proprietary information.

Missing User Warnings

Low

Confidence: 84% confidence
Finding: The README instructs users to export API credentials for Cloudflare and Groq but does not include basic guidance on secret handling, such as avoiding hardcoding, shell history leakage, accidental commits, or sharing logs. While this is documentation rather than executable code, weak secret-handling guidance increases the chance that users expose valid tokens.

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The documented fallback logic routes prompts from a local model to third-party cloud providers, but the skill does not clearly warn that user input may leave the local machine. This is dangerous because users may assume they are using only local/private inference and unknowingly send sensitive prompts, files, or operational context to external services subject to separate retention, logging, and policy controls.

Missing User Warnings

Medium

Confidence: 92% confidence
Finding: The script transmits user prompts to Cloudflare and Groq over the network, but its CLI description and behavior do not clearly warn users that their input may leave the local machine. In a tool marketed as a 'free AI aggregator' with local fallback, this can lead to unintended disclosure of sensitive prompts, especially when users assume local-only execution.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.