Doubao Ai Toolkit

Security checks across malware telemetry and agentic risk

Overview

This is a documentation-only skill for using Doubao/Ark AI CLIs, with expected API-key and media-upload risks but no hidden code or malicious behavior found.

Install only if you trust the npm CLI packages and the Doubao/Ark service. Prefer environment variables or a secret manager for ARK_API_KEY, avoid putting keys directly in commands, and do not send confidential audio, images, video, or text unless your organization permits that provider to process it.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (2)

Missing User Warnings

Medium

Confidence: 97% confidence
Finding: The documentation explicitly shows passing API credentials on the command line and via HTTP headers, which can expose secrets through shell history, process listings, audit logs, terminal recordings, and CI/CD logs. Although the skill also recommends environment variables, providing unsafe credential-handling examples without a clear warning creates a real secret leakage risk in normal use.

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The ASR examples instruct users to upload local audio files and remote audio URLs to an external AI service, but do not warn that potentially sensitive voice content, personal data, or confidential meeting recordings will leave the local environment. In this skill context, ASR is expected behavior, but the absence of a privacy/data-handling warning increases the chance of inadvertent disclosure.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal