Security audit

On Page Seo Checker

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed on-page SEO auditing skill with some scope drift into adjacent SEO/content-quality scoring, but no hidden execution, credential use, exfiltration, destructive behavior, or unsafe persistence was found.

Install this if you want an SEO audit assistant for page-level recommendations. Be aware that its reports may include light E-E-A-T, performance, accessibility, and bulk-audit judgments even though the manifest says those deeper areas belong to other specialist skills; use the dedicated skills for final content-quality, CWV, crawl, indexing, or legal/accessibility conclusions.

A.I.G

Vulnerability Patterns

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
Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
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

Findings (13)

Description-Behavior Mismatch

High

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The document introduces CORE-EEAT profiling, author bio/credentials, citations, first-person narrative, and original data checks even though the skill metadata explicitly says this skill is not for E-E-A-T or publish-readiness scoring. That contradiction can cause the agent to perform disallowed content-quality judgments, misroute work away from the designated content-quality skill, and produce unsupported recommendations with higher policy and quality risk.
Content: No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium

Category: Not specified by scanner
Confidence: 91% confidence
Finding: The manifest description explicitly excludes E-E-A-T and publish-readiness scoring, directing those tasks to a different skill. However, the implementation instructions include Step 10, a "CORE-EEAT Quick Scan," which evaluates 17 items from an E-E-A-T benchmark and therefore extends the audit into an excluded area.
Content: No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium

Category: Not specified by scanner
Confidence: 88% confidence
Finding: A 17-item CORE-EEAT benchmark scan is a distinct content-quality capability beyond the declared scope of structural on-page SEO checks like titles, meta, headers, links, and images. Because the manifest specifically says E-E-A-T / publish-readiness should use another skill, this extra benchmarking capability is context-inappropriate here.
Content: No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium

Category: Not specified by scanner
Confidence: 98% confidence
Finding: The manifest says this skill is 'Not for E-E-A-T / publish-readiness scoring — use content-quality-auditor.' However, the template's step map explicitly includes 'CORE-EEAT quick scan', which adds E-E-A-T-style assessment beyond narrow on-page SEO checks. That is a semantic scope mismatch between the stated skill purpose and the documented behavior.
Content: No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The content-quality template instructs evaluation of 'E-E-A-T signals' alongside depth, proof, and freshness. The manifest explicitly excludes E-E-A-T and publish-readiness scoring, so this documented behavior exceeds the declared scope of an on-page SEO checker.
Content: No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium

Category: Not specified by scanner
Confidence: 89% confidence
Finding: The manifest says the skill is 'Not for crawl / CWV / indexing — use technical-seo-checker.' Yet the template includes mobile, page speed, HTTPS, canonical, and schema markup under 'Technical On-Page', which overlaps with technical SEO concerns the manifest says should be handled elsewhere. This makes the documented behavior broader than the declared intent.
Content: No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium

Category: Not specified by scanner
Confidence: 99% confidence
Finding: This section defines 17 explicit checks spanning intent alignment, citations, author identity, disclosure statements, and first-person experience. Those are classic E-E-A-T/publish-readiness dimensions, which the manifest says this skill should not score. The documentation therefore materially expands the skill's apparent scope.
Content: No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium

Category: Not specified by scanner
Confidence: 98% confidence
Finding: The summary table reserves a scaled 20-point component for 'CORE-EEAT quick scan', and the scaling rule integrates it into the final audit. That means the skill is not merely mentioning E-E-A-T but formally scoring it, directly conflicting with the manifest's exclusion of E-E-A-T/publish-readiness scoring.
Content: No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium

Category: Not specified by scanner
Confidence: 94% confidence
Finding: The playbook expands the skill from its declared single-page on-page audit use case into bulk, site-wide, and pre-publish auditing. This creates scope drift: an orchestrator or user may invoke the skill for tasks it was not approved or described to perform, leading to unreliable outputs, improper delegation, and bypass of intended skill boundaries.
Content: No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The rubric directs the skill to score content quality using E-E-A-T signals even though the manifest explicitly says publish-readiness/E-E-A-T should be handled by a different skill. This creates scope drift that can cause the agent to perform unauthorized or misleading evaluations, reducing reliability and potentially overriding safer routing boundaries between specialized skills.
Content: No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The rubric includes canonical, mobile friendliness, HTTPS, schema, and LCP checks even though the manifest says crawl/CWV/indexing and technical SEO belong to a separate technical-seo-checker skill. This mismatch can cause incorrect tool selection, duplicated responsibilities, and user-facing advice outside the skill's declared competence, weakening control boundaries in the agent system.
Content: No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium

Category: Not specified by scanner
Confidence: 98% confidence
Finding: The dedicated Page Speed section introduces Core Web Vitals-style performance evaluation, which directly conflicts with the manifest statement that crawl/CWV/indexing are out of scope. This is dangerous because it encourages the skill to make performance diagnoses and recommendations that should come from a separate specialist skill, increasing the chance of inaccurate audits and policy bypass through scope expansion.
Content: No source excerpt is available for this finding.

Description-Behavior Mismatch

Low

Category: Not specified by scanner
Confidence: 76% confidence
Finding: The rubric adds a standalone 'Accessibility Overlay' with WCAG 2.2 AA failures and legal-risk framing. While some accessibility checks overlap with SEO, a formal WCAG audit is broader than the manifest's stated focus on titles, meta, headers, keywords, links, and images.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.