Security audit

douyin-downloader-js

Security checks for vulnerabilities and agentic risk

Overview

This Douyin downloader is mostly purpose-aligned, but its network fetching is too broadly scoped and could be abused to reach internal URLs or consume excessive memory/disk.

Install only if you are comfortable running a downloader that fetches remote content and writes files locally. Use a dedicated output directory, avoid untrusted links, and prefer a revised version that restricts hosts to Douyin/CDN domains, limits redirects and file sizes, streams downloads safely, and validates response types.

A.I.G

Vulnerability Patterns

Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
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
Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally

Findings (2)

T09 · Insecure Skill Coding Practices

Error

Location: script/douyin-parse.cjs:38
Finding: Unrestricted HTTP Requests and Redirects Enable Server-Side Request Forgery
Content: View full analysis
{ if (!url.includes('v.douyin.com')) { resolve(url); return; } const parsedUrl = new URL(url); const isHttps = parsedUrl.protocol === 'https:'; const client = isHttps ? https : http; const options = { headers: { 'User-Agent': USER_AGENT }, timeout: 10000 }; const req = client.request(url, options, (res) => { // Check for a redirect if (res.statusCode >= 300 && res.statusCode < 400 && res.headers.location) { console.log('Redirect found:', res.headers.location); // Recursively process the redirect resolveShortUrl(res.headers.location).then(resolve); } else { resolve(res.responseUrl || url); } }); ``` ```js function fetch(url, options = {}) { return new Promise((resolve, reject) => { const parsedUrl = new URL(url); const isHttps = parsedUrl.protocol === 'https:'; const client = isHttps ? https : http; const defaultOptions = { headers: { 'User-Agent': USER_AGENT, 'Referer': REFERER, ...options.headers }, timeout: 30000 }; const req = client.get(url, defaultOptions, (res) => { let data = ''; res.on('data', chunk => data += chunk); res.on('end', () => resolve(data)); }); ``` Media redirects are followed in the same unrestricted manner: ```js const req = client.request(videoUrl, options, (res) => { // Check for a redirect if (res.s ...[truncated 3110 chars]
Remediation: View remediation

T09 · Insecure Skill Coding Practices

Warning

Location: script/douyin-parse.cjs:89
Finding: Unbounded Response Buffering Enables Memory-Exhaustion Denial of Service
Content: View full analysis
{ let data = ''; res.on('data', chunk => data += chunk); res.on('end', () => resolve(data)); }); ``` Image responses are buffered completely before being written: ```js const req = client.request(url, options, (res) => { // Check for a redirect if (res.statusCode >= 300 && res.statusCode < 400 && res.headers.location) { console.log('Image download redirected to:', res.headers.location); downloadImage(res.headers.location, filename).then(resolve).catch(reject); return; } const data = []; res.on('data', chunk => data.push(chunk)); res.on('end', () => { const fs = require('fs'); fs.writeFileSync(filename, Buffer.concat(data)); resolve(filename); }); }); ``` Video responses use the same buffering pattern: ```js const req = client.request(videoUrl, options, (res) => { // Check for a redirect if (res.statusCode >= 300 && res.statusCode < 400 && res.headers.location) { console.log('Video download redirected to:', res.headers.location); // Recursively process the redirect downloadVideoToFile(res.headers.location, filepath).then(resolve).catch(reject); return; } const data = []; res.on('data', chunk => data.push(chunk)); res.on('end', () => { const fs = require('fs'); fs.writeFileSync(filepath, Buffer.concat(data)); resolve(filepath); }); }); ``` ### Technical Analysis The implementation stores entire HTML, image, and video responses in process memory. It does not enforce a maximum byte count or reject an excessive `Content-Length`. For media downloads, `Buffer.concat(data)` c ...[truncated 1710 chars]
Remediation: View remediation

SkillSpector

Vulnerability Patterns

MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
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 (6)

Description-Behavior Mismatch

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The manifest description emphasizes parsing video information and downloading Douyin video/image content, but this file goes beyond retrieval by persisting content to caller-controlled local paths using fs.writeFileSync and creating output directories. Local filesystem modification is a materially broader operation than simple parsing/downloading metadata and is not stated in the manifest text.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 92% confidence
Finding: The code automatically downloads remote video or image content and writes it to local paths under the provided output directory. While there are progress logs, there is no confirmation prompt or clear safety warning before performing these local file writes, which are safety-relevant side effects in a reusable skill context.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: When executed as a script, the program creates an output directory and automatically downloads and saves video content if a media URL is found. Although console logs indicate progress, they do not serve as a prior warning or consent mechanism for the disk write operation.
Content: No source excerpt is available for this finding.

Missing User Warnings

Low

Category: Not specified by scanner
Confidence: 92% confidence
Finding: The skill explicitly downloads remote video or image content into the user's workspace, but it does not clearly warn users that untrusted remote files will be fetched and stored locally. This can lead to accidental ingestion of unsafe or policy-sensitive content, unexpected disk usage, or downstream handling of untrusted media without informed consent.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 84% confidence
Finding: Comments, messages, and user-facing console output are entirely in Chinese, indicating an implicit fixed language choice. There is no indication that the user can opt into another language or that the language restriction is intentionally documented as region-specific.
Content: No source excerpt is available for this finding.

Description-Behavior Mismatch

Low

Category: Not specified by scanner
Confidence: 84% confidence
Finding: When executed directly, the script automatically creates an output directory and saves a video file. The manifest describes a downloader broadly, but does not mention standalone CLI-style execution with implicit local writes, which is a distinct operational behavior.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.