Baidu Qianfan

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed Baidu Qianfan connector that can operate a connected account, including write/delete and billable AI actions, but the behavior is service-scoped and documented.

Install this only if you want an agent to operate Baidu Qianfan through your OOMOL-connected account. Review payloads before approving uploads, create/cancel/delete actions, and remember that generation, batch, OCR, and file workflows may expose submitted content to the provider and consume account credits.

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 (1)

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The description says to use this skill for ANY Baidu Qianfan request and instead of calling the API directly, which is an overly broad invocation trigger. This can cause the agent to route unrelated or higher-risk tasks through a powerful skill with read, create, and delete capabilities, increasing the chance of unintended state-changing operations or misuse without tighter task scoping.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal