Back to skill

Security audit

SEO and LLM Rankings

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent SEO audit tool, but it encourages bypass-style fetching of protected websites and includes unrestricted URL fetching that could reach internal network resources.

Install only if you will use it on sites you own or are authorized to audit. Avoid using its WAF/bot-bypass guidance, do not scan internal or private-network URLs unless explicitly intended, and prefer user-provided HTML or source-code mode when a site blocks automated access.

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 (1)

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/seo_audit.py:77
Finding

Unrestricted URL Fetching Enables Server-Side Request Forgery

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (10)

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
93% confidence
Finding

The skill directs the agent to use network and shell capabilities extensively, but it does not declare an explicit tool scope such as allowed-tools or permissions. In practice, this weakens least-privilege controls and can cause agents or hosts to grant broader access than users expect, especially because the workflow includes arbitrary URL fetching and shell execution.

Content

No source excerpt is available for this finding.

Ssd 2

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill explicitly says certain tools 'bypass WAFs/firewalls' and recommends them as the preferred path for URL fetching. Framing blocked access as a problem to overcome rather than a boundary to respect materially increases the chance that the agent will attempt unauthorized retrieval of protected resources.

Content

No source excerpt is available for this finding.

Ssd 2

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

This section instructs the agent to impersonate a browser User-Agent and use fallback sequencing specifically to get around bot detection and WAF-based blocking. That is dangerous because it encourages indirect circumvention of access controls on third-party sites, which can violate policy and expose the user or platform to legal, operational, or account-abuse consequences.

Content

No source excerpt is available for this finding.

Ssd 2

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The troubleshooting table and notes repeatedly steer the agent toward progressively stronger anti-bot workarounds, including WebFetch preference, browser impersonation, and retry behavior after blocking responses. In context, this turns transient access failures into a playbook for circumvention, increasing the likelihood of unauthorized scraping or fetching from defended sites.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The troubleshooting guidance recommends using curl with --insecure for SSL/TLS failures, which disables certificate validation and enables man-in-the-middle interception or content spoofing. Even though it says to warn the user, the instruction still normalizes an unsafe fallback for a workflow that may influence audit conclusions and downstream code changes.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The HTTP request hard-codes Accept-Language: en-US,en;q=0.9, which imposes an English-language/US locale preference on all audited sites. This is a natural-language locale policy concern because the script does not offer the user any language/locale opt-in or explain why English is required.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The manifest describes a broader capability set that includes prioritized reports with actionable fixes and ready-to-use prompts for AI coding agents. In this file, the implementation only prints console audit sections, issue lists, and scores; aside from a single textual reference to an external template file, it does not actually generate prompts or structured remediation artifacts.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
87% confidence
Finding

The WebPage schema example sets inLanguage to en-US, which is a natural-language locale choice presented as a default template. Because the document does not indicate this is optional or provide alternatives for other locales, it can steer users toward a fixed language/locale without explicit opt-in.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
74% confidence
Finding

The LocalBusiness example uses addressCountry: "US" alongside a US phone format and ZIP field, which imposes a specific locale in natural-language/config content. The file does not clearly state that this locale is merely illustrative or provide guidance for non-US users.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
77% confidence
Finding

The documentation presents the function as performing a basic structural validation, but the implementation applies narrow content requirements such as a top-level heading, blockquote summary, and a Contact section. That is more prescriptive than the docstring suggests and could mislead maintainers about what the validator is actually asserting.

Content

No source excerpt is available for this finding.