Security audit

抖音作品查询

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed Douyin search tool that uses a RedFox API key and offers optional daily subscriptions, with no evidence of hidden or destructive behavior.

Install only if you are comfortable sending Douyin search keywords and date filters to redfox.hk using your REDFOX_API_KEY. Treat the key as sensitive, avoid committing it to files, and review any daily subscription task before allowing it to be created because it may keep running until you remove it.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands

Findings (9)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 83% confidence
Finding: The skill instructs users to configure an API key and invokes an external Python script that likely uses environment variables and network access, but the skill does not declare these capabilities. Undeclared sensitive capabilities reduce transparency and can bypass user expectations or platform policy checks, especially because the script can exfiltrate secrets or make unintended remote requests.

Context-Inappropriate Capability

Medium

Confidence: 88% confidence
Finding: The skill is presented as a search/query tool, but its documentation also directs creation of recurring subscriptions and cron jobs. That expands its behavior from one-time retrieval into persistent execution, which can surprise users, create unauthorized scheduled tasks, and increase the attack surface if abused.

Vague Triggers

Medium

Confidence: 83% confidence
Finding: The README says users can invoke the skill by simply describing the content category they want in natural language, which makes activation boundaries very loose. In an agent environment, broad phrasing increases the chance the skill is triggered during ordinary conversation or ambiguous requests, causing unintended searches, external API use, or subscription flows.

Vague Triggers

Medium

Confidence: 91% confidence
Finding: Examples such as casual conversational search imply the skill may activate on everyday phrases like 'Funny videos seem hot lately,' which overlap heavily with normal chat. That makes accidental invocation more likely, especially in multi-skill systems where generic entertainment or trend-related conversation could be misrouted into this skill.

Missing User Warnings

Low

Confidence: 72% confidence
Finding: The README advertises daily push subscriptions but does not clearly warn that this creates an ongoing notification workflow or explain what user data, identifiers, or retained preferences are involved. This can lead to users consenting without understanding persistence, frequency, or how to stop the notifications.

Vague Triggers

Medium

Confidence: 82% confidence
Finding: The README explicitly says users can invoke the skill with unrestricted natural-language requests, which creates a very broad activation surface and increases the chance the agent will route unrelated or weakly related user input into this skill. In an agent ecosystem, overly broad triggering can cause unintended data access, accidental external API use, or misleading execution when the user did not clearly intend to search Douyin content.

Vague Triggers

Medium

Confidence: 85% confidence
Finding: The example utterances emphasize casual, free-form phrases like trend commentary and implicit requests, which further widens the activation scope beyond precise search commands. This makes accidental invocation more likely and can cause the system to infer intent too aggressively, especially when combined with automatic keyword extraction and subscription behavior.

Vague Triggers

Medium

Confidence: 80% confidence
Finding: The trigger words include broad, everyday terms such as '热门视频' and '抖音搜索', making accidental invocation more likely. Overbroad triggering can cause the skill to run in unrelated contexts, exposing user queries to external services or causing unintended API usage.

Missing User Warnings

Medium

Confidence: 86% confidence
Finding: The documentation asks users to place an API key in a config file or environment variable but does not clearly label it as a sensitive credential or provide handling guidance. This can lead to insecure storage, accidental sharing, or leakage through logs, screenshots, or misconfigured files.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.