ClawHub
Back to skill

Security audit

On Page Seo Auditor

Security checks for vulnerabilities and agentic risk

Overview

This is a markdown-only SEO audit skill with no hidden code or exfiltration, but its audit scope is broader than its stated boundaries.

Install only if you want a broad SEO audit assistant, not a strictly limited on-page checker. Treat E-E-A-T, accessibility, Core Web Vitals, indexing, and large bulk-site conclusions as preliminary and route those areas to specialized tools before acting on recommendations.

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
  • 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 (8)

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The file introduces CORE-EEAT and author-credibility style assessment even though the manifest says this skill is not for E-E-A-T or publish-readiness scoring. That contradiction can override intended task boundaries and push the agent into subjective reputation or quality judgments outside its approved purpose, increasing the risk of misleading audits, policy drift, and unsafe downstream content changes.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The template broadens the skill from narrow on-page SEO into content-quality, E-E-A-T, and technical SEO checks, which conflicts with the manifest-defined scope. In an agent setting, this can cause tool/skill misrouting, produce unsupported evaluations, and lead users to act on low-confidence advice outside the skill’s intended domain.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The file explicitly instructs a 'CORE-EEAT quick scan' even though the manifest says this skill is not for E-E-A-T or publish-readiness scoring. This contradiction is dangerous because agents often follow local templates; the result is unauthorized expansion of analysis scope and potentially misleading judgments presented as part of the audit.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The Content Quality section directs scoring of E-E-A-T signals despite the skill definition excluding E-E-A-T evaluation. This can mislead users into believing the skill provides reliable trust/authority assessments and may contaminate the on-page audit with subjective or unsupported criteria.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The Technical On-Page section includes speed, mobile, and schema checks, which the manifest says should be handled by the technical SEO checker. In practice this creates scope confusion, increases the chance of incomplete or inaccurate technical recommendations, and undermines separation-of-duties between skills.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The playbook explicitly expands the skill from single-page on-page audits to site-wide and bulk auditing, which exceeds the declared manifest scope. This can cause the agent to be invoked for tasks it was not constrained, tested, or approved to perform, leading to inappropriate recommendations, over-collection of user inputs, and unsafe orchestration into other 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 materially expands the skill beyond its declared scope by scoring E-E-A-T, page-speed/mobile LCP, schema, and other technical/accessibility-adjacent factors even though the manifest explicitly says those areas should be handled by other skills. This creates instruction drift: an agent using this rubric may produce misleading audits, over-collect evidence, or make recommendations outside its intended authority, increasing the chance of incorrect or conflicting security-/quality-relevant guidance.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

The accessibility overlay makes WCAG auditing a first-class scored output even though accessibility review is not part of the skill's declared purpose. While not an exploit in the classic sense, this is a genuine scope-integrity issue because it can cause the agent to overstep, generate unsupported compliance conclusions, and blur responsibility between specialized skills.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.