Back to skill

Security audit

Drpy Source Creator

Security checks for vulnerabilities and agentic risk

Overview

The skill mostly matches its drpy source-building purpose, but it needs Review because one helper script has a command-injection bug and some examples can expose playback URLs, cookies, or request details without clear safeguards.

Review the helper scripts before use. Do not run minify_drpy.js with untrusted filenames, avoid hardcoding personal cookies or sharing logs, restrict analyze_site.py to public sites you intend to contact, and only use parser services such as api.apiimg.com if you are comfortable sending playback URLs and related headers to them.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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
Findings (3)

T09 · Insecure Skill Coding Practices

Error
Location
scripts/minify_drpy.js:46
Finding

Shell Command Injection Through User-Controlled File Names

Content
View full analysis
; # ``` The input file must exist, but the attacker-controlled output path does not. ### Impact Assessment Successful exploitation provides arbitrary command execution under the account running the scrip ...[truncated 552 chars]
Remediation
View remediation

T08 · Insecure Dependencies

Warning
Location
references/formatting.md:24
Finding

Unpinned Global Installation of a Third-Party Build Dependency

Content
View full analysis
Remediation
View remediation
``` 2. Commit `package.json` and the generated lockfile. 3. Use `npm ci` in automated and documented workflows to install the locked dependency tree. 4. Invoke the local executable through `npx --no-install uglifyjs` or use the package's Node API. 5. Remove recommendations for global installation. 6. Review package provenance and published integrity data before updating. 7. Use dependency scanning and update the pinned version through a controlled review process. 8. Where supported, disable unnecessary lifecycle scripts during installation and run build tooling in a restricted environment. ]]>

T09 · Insecure Skill Coding Practices

Warning
Location
assets/pitv_example.js:108
Finding

Playback URLs Disclosed to a Hard-Coded Third-Party Parser

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (14)

Tp4

High
Category
MCP Tool Poisoning
Confidence
94% confidence
Finding

The declared description presents a broad drpy source authoring and debugging skill for configuring and troubleshooting video-source definitions. The supplied code instead implements a very specific build utility: it compresses JavaScript files via uglify-js or a simple regex-based fallback. While minification could be loosely related to drpy source handling, it is not described in the declared purpose and is materially different from source configuration, template inheritance, regex assistance, proxy setup, or multi-type source support. The code also uses filesystem access and child-process execution to run an external tool, which are concrete capabilities absent from the description. Therefore this is a meaningful description-behavior mismatch.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
91% confidence
Finding

The skill contains multiple examples and workflows that perform outbound network access (for example, fetching website HTML with Python requests and drpy request-based parsing) but does not declare any tool scope, permissions, or allowed-tools constraints. In an agent environment, undocumented network capability increases the risk of unintended browsing, contacting untrusted domains, or expanding the skill’s effective privileges beyond what reviewers and users expect.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
88% confidence
Finding

The code builds a playback URL using a third-party parsing endpoint and even falls back to a hard-coded external service when no parser URL is found. This can transmit user-requested media URLs and related context to an untrusted external domain, creating privacy, tracking, integrity, and supply-chain risks; in this skill context, the danger is higher because the skill is specifically designed to generate and debug scraping/playback rules that routinely process user-selected content URLs.

Content

Scanner excerpt · assets/pitv_example.js (reported line 112)May include surrounding context.

js
if (apiMatch) {
        parseApi = apiMatch[1];
      } else {
        parseApi = 'https://api.apiimg.com/show/super.php?id=';
      }
      if (parseApi.startsWith('//')) {
        parseApi = 'https:' + parseApi;

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The document explicitly teaches use of outbound requests, local proxy handlers, sniffer rules, dynamic JavaScript hooks, and custom headers/cookies, but it does not warn users about privacy leakage, unsafe request routing, local service exposure, or the risk of executing untrusted scraping logic. In a skill meant to create and debug drpy sources, these capabilities are expected, but the omission of security guidance can lead users to build sources that unintentionally expose local resources, leak credentials, or fetch attacker-controlled content.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The guide explicitly suggests placing a raw Cookie value into request headers to bypass 403-style protections, but gives no warning that cookies may contain live session tokens or user-auth data. In this skill context, users are debugging third-party scraping/video sources, which increases the chance they will copy browser session cookies into configs that may later be shared, logged, or reused beyond their intended scope.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The logging snippet wraps all requests and logs full URLs, request options, response lengths, and errors without any sanitization guidance. In this skill’s debugging context, request options may include Cookie, Referer, tokens, or signed media URLs, so the advice can easily cause credential leakage into logs, screenshots, shared bug reports, or persistent local files.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The script performs an outbound HTTP request directly to a user-supplied URL without any validation, allowlisting, or explicit warning that network access will occur. This creates an SSRF-style risk in environments where the tool may be run against attacker-controlled input, potentially reaching internal services, cloud metadata endpoints, or other sensitive network locations.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

After parsing the target site, the script automatically issues a second request to a discovered search endpoint (url + action) without clearly disclosing this additional probe. That expands the attack surface beyond the originally supplied URL and can trigger unintended requests to attacker-influenced paths, including malformed or sensitive internal endpoints if the analyzed page content is malicious.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

The file’s natural-language comments and usage instructions are entirely in Chinese, including the template description and setup guidance. Under the stated policy, forcing a specific language without opt-in can be a locale/language policy violation when no alternative or user choice is provided.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
87% confidence
Finding

The instructional and descriptive text in comments is entirely Chinese, which may impose a language choice on users or maintainers without opt-in. The file does not state that it is intended only for a Chinese-speaking or region-specific audience.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
80% confidence
Finding

The code makes an HTTP request via request(this.homeUrl) to fetch remote content. While the file is an example scraper template and includes technical comments, it does not provide a user-facing warning, confirmation, or explicit disclosure near the operation that the skill will contact external sites.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
86% confidence
Finding

The file's comments, field labels, and user-facing descriptors are all in Chinese, and there is no indication that users can opt into another language or that the skill is intentionally limited to a Chinese-speaking audience. Under the language/locale policy, forcing a specific language without user opt-in can be a natural-language policy violation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

The document title and all instructional content are written exclusively in Chinese, with no indication that other languages are supported or that the user can choose a locale. Under the stated policy, forcing a single language without opt-in is a natural-language locale violation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

This code file contains all user-facing console messages in Chinese, including usage, errors, warnings, and summary text. Because the script does not offer a language option or document that it is intentionally limited to a Chinese-speaking audience, it forces a specific language without user opt-in.

Content

No source excerpt is available for this finding.