Security audit

Comfyui anfrage

Security checks across malware telemetry and agentic risk

Overview

This skill is a straightforward ComfyUI client that sends user-provided workflows to a configured ComfyUI server, with network and credential risks users should understand.

Install this only if you intend to send workflows, prompts, and generated image metadata to a ComfyUI server you control or trust. Set COMFYUI_HOST and COMFYUI_PORT deliberately, avoid reusing important passwords for COMFYUI_USER/COMFYUI_PASS, and prefer localhost, a trusted LAN/VPN, or an HTTPS reverse proxy when prompts or credentials are sensitive.

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: 94% confidence
Finding: The skill declares no explicit permissions even though its documented behavior requires network access and reads environment variables for host and credentials. This weakens transparency and policy enforcement, making it easier for a caller to invoke outbound requests or credential-dependent behavior without clear user awareness or review.

Missing User Warnings

Medium

Confidence: 96% confidence
Finding: The skill documentation says it POSTs workflow data to an HTTP endpoint but does not warn that the contents are transmitted in cleartext. If used on an untrusted network, workflows and server responses could be intercepted or modified, which is especially risky if prompts, embedded data, or generated image links are sensitive.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The skill instructs users to supply basic-auth credentials via environment variables but does not explain handling, storage, or exposure risks. Credentials may be sent over plain HTTP, appear in logs or debug output, or be reused insecurely, increasing the chance of disclosure and unauthorized access to the ComfyUI service.

Missing User Warnings

Medium

Confidence: 92% confidence
Finding: The CLI sends the supplied workflow JSON to a remote ComfyUI server over the network without any explicit user disclosure, confirmation, or indication of destination sensitivity. In this skill, workflows may contain prompts, embedded metadata, file references, or other sensitive content, so silent transmission can expose user data unexpectedly; the risk is increased because the default target is a non-local private IP and the transport uses plain HTTP rather than HTTPS.

Missing User Warnings

Medium

Confidence: 90% confidence
Finding: The code automatically attaches Basic Authentication credentials from environment variables to outbound requests without notifying the user. This is risky because credentials may be sent to an unexpected or misconfigured host, and in this implementation they are sent over HTTP by default, making interception or credential disclosure more likely on untrusted networks.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

Detected: suspicious.env_credential_access

Environment variable access combined with network send.

Critical

Code: suspicious.env_credential_access
Location: bin/cli.js:5