Back to skill

Security audit

抖音视频文案智能提取

Security checks for vulnerabilities and agentic risk

Overview

The skill appears intended to transcribe and rewrite videos through an external service, but its URL download and upload behavior is broader and less clearly bounded than its documentation claims.

Review before installing. Only use this skill with videos and URLs you are comfortable sending to the Guaikei service and its storage provider. Avoid running it on confidential local media or on URLs from untrusted sources, especially internal, localhost, private-network, or metadata-service URLs. Treat the deletion and no-training statements as provider promises rather than behavior proven by the client code.

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 (3)

T09 · Insecure Skill Coding Practices

Error
Location
scripts/utils/validator.js:3
Finding

Arbitrary URL Fetching Enables Server-Side Request Forgery and Internal Data Exfiltration

Content
View full analysis
{ this.__response = response; if (!this.__isResumed) { this.__total = parseInt(response.headers["content-length"]) || null; this.__resetStats(); } if (this.__isResumed && response.statusCode === 200) { this.__isResumed = false; this.__total = parseInt(response.headers["content-length"]) || null; this.__resetStats(); } if (this.__isRequireRedirect(response)) { this.__redirectCount++; if (this.__redirectCount > this.__opts.maxRedirects) { const err = new Error("Too many redirects"); this.__setState(this.__states.FAILED); this.emit("error", err); return reject(err); } const redirectedURL = /^https?:\/\//.test(response.headers.location) ? response.he ...[truncated 3288 chars]
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/utils/helper.js:20
Finding

Unbounded Remote Downloads Can Exhaust Disk, Bandwidth, and Execution Time

Content
View full analysis
{ const options = { retry: { maxRetries: constants.CREATE_MAX_ATTEMPTS, delay: constants.RETRY_INTERVAL, }, override: { skip: true, skipSmaller: true }, }; let progressLog = ""; const dl = new Downloader(url, path, options); dl.on("end", (info) => { inlineLog(progressLog + " - Download completed"); resolve(info); }) .on("skip", (skipInfo) => { inlineLog(`File ${skipInfo.fileName} already exists; skipping`); resolve(skipInfo); }) .on("retry", (attempt, opts) => { let count = Math.floor(opts.delay / 1000); const retryLog = () => { inlineLog(`Retry ${attempt}/${opts.maxRetries}; waiting ${count} seconds`); if (count > 0) { setTimeout(() => retryLog(), 1000); } count--; }; retryLog(); }) .on("resume", (isResumed) => { if (!isResumed) { inlineLog("This URL does not support resume; the file will restart."); } }) .on("progress", (stats) => { const progress = stats.progress.toFixed(1); const speed = byteHumanize(stats.speed); const downloaded = byteHumanize(stats.downloaded); const total = byteHumanize(stats.total); progressLog = `${stats.name}: ${speed}/s - ${progress}% [${downloaded}/${total}]`; inlineLog(progressLog); }) .on("error", (error) => { inlineLog("Download failed:" + error.message); reject(error); }); dl.start().catch((err) => { in ...[truncated 3576 chars]
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/utils/upload.js:12
Finding

Server-Controlled Presigned Upload Destination and Headers Are Insufficiently Restricted

Content
View full analysis
{ const fileStats = fs.statSync(filename); const totalSize = fileStats.size; let uploadedSize = 0; let settled = false; const fileStream = fs.createReadStream(filename); const uploadHeaders = Object.assign({}, headers); uploadHeaders["Content-Length"] = String(totalSize); if (!uploadHeaders["Content-Type"]) { uploadHeaders["Content-Type"] = "application/octet-stream"; } const options = { host: url.hostname, path: url.pathname + url.search, method: "PUT", headers: uploadHeaders, }; const req = https.request( { ...options, timeout: constants.REQUEST_TIMEOUT }, (res) => { ``` ```md
Remediation
View remediation
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (24)

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The manifest heavily emphasizes obtaining and using a private third-party token and includes out-of-band contact details to acquire access, while the core declared media-processing behavior is not machine-verifiable from the file. In a skill ecosystem, this creates supply-chain and data-governance risk: users may grant secrets and upload media to an external service with limited enforceable guarantees beyond prose claims.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The manifest heavily emphasizes obtaining and using a private third-party token and includes out-of-band contact details to acquire access, while the core declared media-processing behavior is not machine-verifiable from the file. In a skill ecosystem, this creates supply-chain and data-governance risk: users may grant secrets and upload media to an external service with limited enforceable guarantees beyond prose claims.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

The manifest heavily emphasizes obtaining and using a private third-party token and includes out-of-band contact details to acquire access, while the core declared media-processing behavior is not machine-verifiable from the file. In a skill ecosystem, this creates supply-chain and data-governance risk: users may grant secrets and upload media to an external service with limited enforceable guarantees beyond prose claims.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The manifest heavily emphasizes obtaining and using a private third-party token and includes out-of-band contact details to acquire access, while the core declared media-processing behavior is not machine-verifiable from the file. In a skill ecosystem, this creates supply-chain and data-governance risk: users may grant secrets and upload media to an external service with limited enforceable guarantees beyond prose claims.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The manifest heavily emphasizes obtaining and using a private third-party token and includes out-of-band contact details to acquire access, while the core declared media-processing behavior is not machine-verifiable from the file. In a skill ecosystem, this creates supply-chain and data-governance risk: users may grant secrets and upload media to an external service with limited enforceable guarantees beyond prose claims.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The manifest heavily emphasizes obtaining and using a private third-party token and includes out-of-band contact details to acquire access, while the core declared media-processing behavior is not machine-verifiable from the file. In a skill ecosystem, this creates supply-chain and data-governance risk: users may grant secrets and upload media to an external service with limited enforceable guarantees beyond prose claims.

Content

No source excerpt is available for this finding.

Whitespace Padding

Medium
Category
Prompt Injection
Confidence
70% confidence
Finding

Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Content

Scanner excerpt · README.md (reported line 99)May include surrounding context.

md
4. 同时传入文件路径与任务ID,优先执行 `--id`,忽略 `--file`
5. 无自定义 prompt 时,默认完整转录视频全部文字

| 用户自然语言指令                                         | 自动生成命令                                                                                                |
| -------------------------------------------------------- | ----------------------------------------------------------------------------------------------------------- |
| 视频提取 https://example.com/video.mp4 中的文字          | `node scripts/video2text/index.js --file "https://example.com/video.mp4"`                                   |
| 把本地 /path/to/your/video.mp4 改成小红书风格的文案      | `node scripts/video2text/index.js --file "/path/to/your/video.mp4" --prompt "改写成小红书风格的文案"`       |

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
79% confidence
Finding

The skill declares access to a sensitive environment variable (GUAIKEI_API_TOKEN) but does not define an explicit tool/permission scope such as permissions or allowed-tools. That weakens least-privilege guarantees and makes it harder for a host agent to constrain what the skill may access or expose, especially since the skill also handles user-supplied URLs and local file paths.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
87% confidence
Finding

The invocation guidance is intentionally broad and encourages routing many generic content-analysis, meeting-summary, rewriting, and translation requests through this skill, not just Douyin-specific extraction. In an agent setting, overbroad triggers can cause unintended transmission of arbitrary user media or local files to the external service, increasing privacy and exfiltration risk beyond what the skill name suggests.

Content

No source excerpt is available for this finding.

Whitespace Padding

Medium
Category
Prompt Injection
Confidence
70% confidence
Finding

Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Content

Scanner excerpt · SKILL.md (reported line 120)May include surrounding context.

md
当用户用自然语言下达指令时,按以下映射生成命令,保证识别与执行一致:

| 用户自然语言指令                                     | 生成的命令                                                                                                  |
| ---------------------------------------------------- | ----------------------------------------------------------------------------------------------------------- |
| 提取 https://example.com/video.mp4 里的文字          | `node scripts/video2text/index.js --file "https://example.com/video.mp4"`                                   |
| 总结这个视频的核心观点 https://example.com/video.mp4 | `node scripts/video2text/index.js --file "https://example.com/video.mp4" --prompt "总结这个视频的核心观点"` |

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The natural-language content in this file, including function descriptions and user-visible error messages, is entirely in Chinese with no indication that language selection is configurable or intentionally limited to a Chinese-only context. This can violate language/locale policy when a skill imposes a specific language without offering user choice or documenting the constraint.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

This JavaScript file contains user-facing comments and help/error text in Chinese, starting with the file description and continuing throughout the CLI parser. Under the policy, forcing a specific language without user opt-in or documented justification is a natural-language policy violation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The thrown errors and generated help text are all hardcoded in Chinese, including missing-value errors, usage text, option descriptions, and warnings. Because the file provides user-facing natural-language output but no language selection or documented locale restriction, it violates the language/locale policy.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
83% confidence
Finding

This code issues HTTP/HTTPS requests to arbitrary URLs and writes the response to the local filesystem, then deletes the downloaded file automatically when stop() is called if removeOnStop is enabled. Although the behavior is functional for a downloader, there is no visible confirmation prompt or user-facing disclosure in this file about the destructive deletion behavior or remote data retrieval.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The __removeFile() path closes the file stream and unlinks the downloaded file whenever removeOnFail is true, which is enabled by default in the options. This is a potentially destructive operation, but the file contains no explicit warning, confirmation, or user-visible disclosure about automatic cleanup of partial downloads on errors/timeouts.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

This code includes natural-language comments and all user-visible status/error messages exclusively in Chinese, such as the docstring and download progress strings. For a general-purpose helper module, this imposes a specific language on users without opt-in or explanation, which matches the language/locale policy violation criteria.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

This code emits all user-visible error messages in Chinese, including operational guidance and remediation text. That creates a language policy concern because it imposes a specific language on users without offering a choice or documenting a justified locale restriction.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The file emits all user-facing warning and informational strings exclusively in Chinese, including the key error path when the token is invalid. This imposes a specific language on users without any opt-in, fallback, or indication that the skill is intentionally limited to a Chinese-speaking or region-specific context.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The comment makes a privacy and data-handling promise that uploaded videos are only used for transcription and automatically deleted afterward, but this function only performs an upload and contains no deletion, retention enforcement, or verification mechanism. In a skill that processes user videos, this creates a real risk of misleading users and retaining potentially sensitive media longer than expected, which can cause privacy, compliance, and trust issues.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The banner text is fixed in Chinese ("视频文案智能提取助手"), which imposes a specific language on users. Under the policy, locale or language should not be forced without opt-in unless clearly documented and justified.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

The README instructs users to obtain a token via WeChat and presents the setup flow entirely in Chinese, with no indication that other languages are supported or that the skill is intentionally restricted to a Chinese-speaking region. This can constitute a language/locale policy issue because it imposes a specific language channel on users without opt-in.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
78% confidence
Finding

The manifest description is entirely in Chinese and presents the skill's behavior only in that language. For a general-purpose skill manifest, this can create a locale policy concern because users are not offered an alternative language or an explicit region-specific justification.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
80% confidence
Finding

This code sends user-provided video analysis data, including the video URL and optional prompt, to a remote API via requestApi. While the functions have developer-facing comments, there is no visible confirmation prompt, log, or other user-facing disclosure in this file indicating that user content will be transmitted off-box for analysis.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
93% confidence
Finding

This file contains natural-language comments and error strings exclusively in Chinese, including operational messages shown to users such as upload failures and protocol requirements. Under the stated policy, forcing a specific language without user opt-in can be a locale-policy violation when no justification or language selection mechanism is present.

Content

No source excerpt is available for this finding.