Security audit

AdsTurbo 广告视频复刻 · AdsTurbo Ad Video Clone

Security checks for vulnerabilities and agentic risk

Overview

This skill performs the advertised AdsTurbo ad-video analysis and generation workflow, with disclosed uploads and API use, but users should avoid uploading private or unauthorized media.

Install only if you intend to use AdsTurbo for ad-video analysis or cloning. Do not upload private, confidential, regulated, copyrighted, or likeness-sensitive media unless you have the rights and consent to use it and are comfortable with it being sent to AdsTurbo and made available by URL.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (7)

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The trigger phrases include very broad terms such as "replicate it" and "analyze this ad," which can match many ordinary user requests and cause this skill to activate outside its intended narrow context. Because this skill enables cloning a competitor ad and sending user-provided video URLs to an external service, accidental over-activation can route unrelated requests into a higher-risk workflow involving IP-sensitive imitation and third-party data transfer.

Vague Triggers

Medium

Confidence: 87% confidence
Finding: The Chinese trigger guidance uses similarly broad phrases like 「仿制」 and 「分析这条广告」 that can fire on ambiguous requests, increasing the chance of unintended invocation for general analysis or creative-assistance queries. In this skill's context, mistaken activation is more concerning because the workflow is specifically designed to clone ad structure and interact with an external API, amplifying policy, copyright, and privacy risks if triggered too loosely.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The skill instructs users to upload reference videos to public URLs before analysis or cloning, but provides no warning about privacy, confidentiality, or intellectual-property exposure. This can lead users to unintentionally disclose sensitive or proprietary media, and because the content must be publicly accessible, the risk is heightened beyond normal file upload behavior.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The skill promotes cloning a video's style and chaining into character-swap or translation workflows, but omits warnings about copyright, likeness, identity, and consent risks. In this context, users may reasonably infer that cloning competitor ads or swapping people into cloned videos is an endorsed use, increasing the chance of misuse for infringement or deceptive synthetic media.

Missing User Warnings

Medium

Confidence: 92% confidence
Finding: The instructions direct users to upload local assets to a publicly accessible URL-backed service, but they do not warn that private files will be transmitted off-device and made reachable via a public URL. This can lead users to expose sensitive photos, audio, or video unintentionally, especially because the document frames upload as a mandatory prerequisite for normal operation.

Missing User Warnings

Medium

Confidence: 92% confidence
Finding: The Chinese section repeats the same upload workflow without any privacy or public-accessibility warning, so Chinese-reading users receive operational guidance but not the associated data-exposure risk. Because the file emphasizes that local files must be uploaded first, users may reasonably assume this is safe by default and disclose sensitive media inadvertently.

Unpinned Dependencies

Low

Category: Supply Chain
Content: requests>=2.28.0
Confidence: 97% confidence
Finding: The dependency is specified with a lower-bound constraint only (`requests>=2.28.0`), which makes builds non-reproducible and allows unexpected package versions to be installed over time. This increases supply-chain risk and can inadvertently pull in vulnerable or breaking releases depending on when installation occurs.

Static analysis

No suspicious patterns detected.