Security audit

You Get

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent media-downloader guide, but it needs Review because it includes under-scoped instructions involving browser cookies, unpinned installs, and disabling the system firewall.

Before installing, review every command the agent proposes. Do not provide a browser cookie database or exported cookies unless you understand they can act like account credentials, avoid the mutable develop-branch install, use an isolated Python environment, confirm output paths before downloads, and do not run the firewall-disable troubleshooting step.

A.I.G

Vulnerability Patterns

Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Insecure DependenciesIntroduces malicious components through unsafe dependency sources
Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review

Findings (2)

T05 · Unauthorized Access and Privilege Escalation

Error

Location: troubleshooting.md:360
Finding: System-Wide Firewall Disabled During Network Troubleshooting
Content: View full analysis
Remediation: View remediation

T08 · Insecure Dependencies

Warning

Location: troubleshooting.md:199
Finding: Installation of Mutable and Unpinned Third-Party Dependencies
Content: View full analysis
Remediation: View remediation

SkillSpector

Vulnerability Patterns

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
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
YARA SignaturesMalware Match, Webshell Match, Cryptominer Match

Findings (25)

Missing User Warnings

High

Category: Not specified by scanner
Confidence: 97% confidence
Finding: Advertising cookie-based authenticated downloads without any credential-handling guidance is risky because browser cookies can grant access to private accounts and paid or personal content. In an AI-agent workflow, users may be encouraged to expose sensitive session material without understanding retention, logging, scope, or the consequences of cookie theft or misuse.
Content: No source excerpt is available for this finding.

Credential Access

High

Category: Privilege Escalation
Confidence: 97% confidence
Finding: The explicit instruction to export cookies from Firefox or Chrome describes obtaining authentication material from browser stores. In an agent context, this is credential access behavior because those cookies may grant account access and can be repurposed outside the intended download session.
Content: Scanner excerpt · guides/03-advanced-usage.md (reported line 79)May include surrounding context.
md
**步骤**: 1. 使用浏览器插件导出cookies.txt 2. 或从Firefox/Chrome的cookies.sqlite导出 ### 使用Cookies下载

Credential Access

High

Category: Privilege Escalation
Confidence: 98% confidence
Finding: Pointing the tool at a Firefox cookies.sqlite path directs the agent to read a browser credential database directly. This is especially dangerous because it operationalizes local credential harvesting and can expose active sessions far beyond the specific media site in question.
Content: Scanner excerpt · guides/03-advanced-usage.md (reported line 88)May include surrounding context.
you-get -c cookies.txt 'https://www.youtube.com/watch?v=xxxxx'

使用Firefox cookies数据库

you-get -c ~/.mozilla/firefox/xxxxx/cookies.sqlite 'URL'

text
---

YARA rule 'agent_skill_remote_bootstrap_execution': Remote script or code download followed by execution/bootstrap installation [agent_skills]

High

Category: YARA Match
Confidence: 90% confidence
Finding: The document recommends installing directly from a GitHub develop branch via pip install --upgrade git+https://..., which pulls and executes unpinned remote code outside the vetted package release channel. This increases supply-chain risk because users may install unstable or compromised code with no version pinning, integrity verification, or trust boundary explanation.
Content: Scanner excerpt · troubleshooting.md (reported line 200)May include surrounding context.
/www.bilibili.com/video/BV1xx411c7mD'

2. 检查是否地区限制

curl -I 'https://www.bilibili.com'

3. 尝试更新you-get

pip install --upgrade you-get

text
**常见原因**: - Bilibili更新反爬机制 (40%) - 视频需要登录 (30%) - 高画质需要会员 (20%) - you-get版本过旧 (10%) **解决方案**: **方案A（更新）**: 更新you-get ```bash pip install --upgrade you-get pip install --upgrade git+https://github.com/soimort/you-get@develop

方案B（登录）: 使用cookies

bash
you-get -c cookies.txt 'URL'

方案C（备用）: 使用其他工具

bash
# 如you-get不支持，可尝试youtube-dl或yt-dlp pip install yt-dlp yt-dlp 'URL'

格式与质量问题

6. 无法下载1080p视频【简单问题】

问题描述: 只能下载720p或更低分辨率

排查步骤:

bash
# 查看所有可用格式 you-get -i 'URL'

常见原因:

1080p+需要FFmpeg合并 (60%)

视频本身最高只有720p (30%)

格

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 91% confidence
Finding: The skill explicitly states the AI will execute download commands and monitor progress, but it does not disclose that these actions write files locally, consume bandwidth and storage, and may place content into default or user-specified directories. In an agent setting, omission of these side effects can cause unexpected disk usage or accidental overwriting of files, especially when handling large media downloads.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 94% confidence
Finding: The guide explicitly tells an AI it can directly run package installation commands (pip install you-get) without any approval gate or warning that this will modify the local system. In an agentic context, converting documentation into executable package-management actions increases the risk of unauthorized software installation, dependency confusion, or environment modification on the user's machine.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: This section includes privileged package-install instructions, including sudo apt install ffmpeg and sudo yum install ffmpeg, without any warning about elevated privileges or system-wide changes. In an AI-execution setting, this is more dangerous than ordinary documentation because it can cause an agent to perform root-level modifications automatically, expanding impact if the command is misapplied or the environment is sensitive.
Content: No source excerpt is available for this finding.

Sudo/Root Execution

Medium

Category: Privilege Escalation
Confidence: 95% confidence
Finding: The document contains a sudo command that installs software with root privileges. While common in admin documentation, its inclusion in a skill that instructs AI to perform setup makes it hazardous because an agent may execute privileged commands without sufficient validation, user consent, or environment checks.
Content: Scanner excerpt · guides/01-installation.md (reported line 94)May include surrounding context.
bash
# Ubuntu/Debian sudo apt install ffmpeg # CentOS/RHEL sudo yum install ffmpeg

Sudo/Root Execution

Medium

Category: Privilege Escalation
Confidence: 95% confidence
Finding: This sudo yum install ffmpeg command grants root-level package installation capability and can materially alter the host system. In the context of an agent skill, the risk is amplified because the skill frames environment setup as something the AI may carry out, making unauthorized privileged execution a realistic failure mode.
Content: Scanner excerpt · guides/01-installation.md (reported line 97)May include surrounding context.
sudo apt install ffmpeg

CentOS/RHEL

sudo yum install ffmpeg

text
### Windows安装

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 94% confidence
Finding: The guide explicitly tells an AI to directly execute a network download command, which causes outbound network access and writes downloaded content to local storage without any user-consent, destination-review, or safety warning. In an agent context, this can lead to unintended retrieval of untrusted remote content and silent filesystem changes, making the skill materially more dangerous than ordinary human-facing CLI documentation.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The save-path section authorizes AI-directed writes to user-selected locations, including arbitrary directories and filenames, without guardrails about overwriting, sensitive paths, disk usage, or path validation. In an agent setting, this increases risk because the model may write into important user directories or unexpected locations based on ambiguous prompts.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The force-redownload example uses an overwrite behavior and notes that existing files may be replaced, but it does not require confirmation or explain the risk of destructive file loss. In an AI-agent workflow, this is dangerous because an autonomous retry or user-misinterpreted instruction could clobber previously saved content without notice.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 91% confidence
Finding: Natural-language policy violations include forcing a specific language without user opt-in. This file presents all instructional content in Chinese and does not indicate that the user can select another language or that the guide is intentionally limited to a Chinese-speaking or region-specific audience.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The documentation instructs use of exported browser cookies and direct browser cookie databases for authenticated downloads without any sensitive-data handling guidance. Cookies.sqlite and cookies.txt can contain live session credentials, so encouraging agent access to them risks credential theft, privacy violations, and unauthorized account use.
Content: No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium

Category: Not specified by scanner
Confidence: 89% confidence
Finding: The guide instructs creation of shell scripts and parallel command execution, moving beyond simple downloader usage into general automation. In an agent environment, generating files and running loops or xargs can amplify unintended network activity, persistence, and command misuse if inputs are not tightly controlled.
Content: No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The guide expands a media-downloading skill into arbitrary file retrieval by explicitly documenting direct download of any URL such as a PDF. In an agent context, this broadens capability from a bounded media tool to generic network file acquisition, which can be abused to fetch untrusted payloads or sensitive content without additional safeguards.
Content: No source excerpt is available for this finding.

Sudo/Root Execution

Medium

Category: Privilege Escalation
Confidence: 70% confidence
Finding: Commands invoke sudo or root privileges. Verify this elevated access is necessary and justified.
Content: Scanner excerpt · troubleshooting.md (reported line 74)May include surrounding context.
Ubuntu/Debian:

bash
sudo apt install ffmpeg

Windows:

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The troubleshooting guide instructs users to disable the firewall with sudo ufw disable as a test step, but provides no warning about the security implications or safer alternatives. This can expose the host to unnecessary network risk, especially if users forget to re-enable the firewall or run the command on an internet-connected system.
Content: No source excerpt is available for this finding.

Sudo/Root Execution

Medium

Category: Privilege Escalation
Confidence: 97% confidence
Finding: This step tells users to run a privileged command that disables the Linux firewall, materially reducing host protections during troubleshooting. Because it is presented as a simple test without prominent risk warnings or recovery guidance, it may lead users to weaken system security unnecessarily.
Content: Scanner excerpt · troubleshooting.md (reported line 362)May include surrounding context.
方案B（检查防火墙）:

bash
# 临时关闭防火墙测试 sudo ufw disable # Linux

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 80% confidence
Finding: The natural-language description and headings are entirely Chinese, which can amount to a language/locale constraint if the skill is presented without user opt-in or an explicit scope limitation. There is no indication that the skill supports multiple languages or is intentionally limited to a Chinese-speaking audience for a documented reason.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 90% confidence
Finding: Natural-language comments and headings throughout the file are presented exclusively in Chinese, with no indication that the skill is intended only for a Chinese-speaking or region-specific audience. Under the stated policy, forcing a specific language without opt-in can be a locale policy violation.
Content: No source excerpt is available for this finding.

Missing User Warnings

Low

Category: Not specified by scanner
Confidence: 82% confidence
Finding: This shell script contains multiple example commands that make network requests to third-party sites and, in several cases, save downloaded media to the local system. Although these are examples, the file provides no user-facing warning in comments about network access or local file creation, which is relevant safety context for a code file.
Content: No source excerpt is available for this finding.

Context-Inappropriate Capability

Low

Category: Not specified by scanner
Confidence: 86% confidence
Finding: The guide authorizes launching local external applications such as VLC, mpv, or Chromium. While common for media playback, this creates a boundary-crossing action for an agent by invoking local programs that may process untrusted remote content and trigger additional side effects.
Content: No source excerpt is available for this finding.

Missing User Warnings

Low

Category: Not specified by scanner
Confidence: 82% confidence
Finding: The guide recommends creating and executing a batch download script without warning about writing files or repeated outbound requests. In an agent setting, this can unintentionally produce persistent artifacts and excessive network activity, especially if the URL list is attacker-influenced.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 84% confidence
Finding: The skill content forces a specific language/locale for all instructions without indicating user opt-in or that the skill is specifically intended for Chinese-speaking users. The policy allows locale constraints when explicitly documented and justified, but this file provides no such note.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.

Overview

A.I.G

T05 · Unauthorized Access and Privilege Escalation

T08 · Insecure Dependencies

SkillSpector

Missing User Warnings

Credential Access

Credential Access

使用Firefox cookies数据库

YARA rule 'agent_skill_remote_bootstrap_execution': Remote script or code download followed by execution/bootstrap installation [agent_skills]

2. 检查是否地区限制

3. 尝试更新you-get

格式与质量问题

6. 无法下载1080p视频【简单问题】

Missing User Warnings

Missing User Warnings

Missing User Warnings

Sudo/Root Execution

Sudo/Root Execution

CentOS/RHEL

Missing User Warnings

Missing User Warnings

Missing User Warnings

Natural-Language Policy Violations

Missing User Warnings

Context-Inappropriate Capability

Description-Behavior Mismatch

Sudo/Root Execution

Missing User Warnings

Sudo/Root Execution

Natural-Language Policy Violations

Natural-Language Policy Violations

Missing User Warnings

Context-Inappropriate Capability

Missing User Warnings

Natural-Language Policy Violations

Static analysis