Security audit

Jianying Video Gen

Security checks across malware telemetry and agentic risk

Overview

This skill appears to generate Jianying videos as advertised, but it ships and automatically uses real-looking browser session cookies and has an undisclosed output-path mismatch.

Review before installing. Delete the bundled cookies.json, do not rely on or share packaged session cookies, and provide only your own Jianying credentials if you trust the publisher. Treat cookies.json like a password, expect prompts and selected media to be uploaded to Jianying, confirm credit costs before generation, and set an explicit output_dir to avoid unexpected local writes.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (7)

Context-Inappropriate Capability

High

Confidence: 99% confidence
Finding: The file contains live-looking authenticated Jianying session cookies and related tokens embedded directly in the skill package. Hardcoded session credentials are sensitive secrets that can allow anyone with access to the skill package to impersonate the account, generate content, access account data, or perform actions as the user; in this skill’s context, the inclusion is especially dangerous because the skill explicitly requires cookies for login and therefore bundles operational account access, not just code.

Description-Behavior Mismatch

Medium

Confidence: 80% confidence
Finding: The tool schema says `output_dir` defaults to the current directory, but runtime behavior silently writes to `D:\SQLMessage\AI_Videos` when the user omits it. This mismatch can cause generated files to be written to an unexpected local location, which is a security-relevant trust and transparency issue, especially for tools that create local artifacts.

Intent-Code Divergence

Medium

Confidence: 80% confidence
Finding: The documented default output directory conflicts with the effective runtime default, creating misleading behavior around where files are stored. In a skill that generates and writes media files, undisclosed output locations can lead to accidental exposure, confusion, or policy violations on the host machine.

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The documentation instructs users to export and supply login cookies for a live account but does not warn that these cookies are highly sensitive session credentials that can enable account takeover if exposed. In this skill's context, the risk is elevated because the automation is designed to authenticate to a third-party service and users may casually store or share cookies.json without understanding the consequences.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The skill sends user prompts and optionally local images/videos to xyq.jianying.com, but the documentation does not clearly warn that this content is transmitted to a third-party service. This omission can cause users to upload sensitive or proprietary material without informed consent, creating privacy, confidentiality, and compliance risks.

Missing User Warnings

Medium

Confidence: 74% confidence
Finding: The skill creates directories and writes generated video output to a user-specified path without any explicit disclosure in the execution flow shown here. In context, file creation is expected for a video-generation skill, so the danger is limited, but the lack of clear notice can still surprise users and cause unintended writes to sensitive or monitored locations.

Missing User Warnings

Medium

Confidence: 83% confidence
Finding: The skill requires and accesses `cookies.json`, which contains login credential material, yet this code path provides no clear user-facing warning that account session credentials will be consumed by the worker. Because these cookies likely grant access to a third-party account, undisclosed credential use increases the risk of user surprise, misuse, and accidental exposure through the surrounding toolchain.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.