Trust Assessment
seo-audit received a trust score of 85/100, placing it in the Mostly Trusted category. This skill has passed most security checks with only minor considerations noted.
SkillShield's automated analysis identified 1 finding: 0 critical, 1 high, 0 medium, and 0 low severity. Key findings include Potential Data Exfiltration via Reading User Context File.
The analysis covered 4 layers: dependency_graph, manifest_analysis, llm_behavioral_safety, static_code_analysis. All layers scored 70 or above, reflecting consistent security practices.
Last analyzed on February 16, 2026 (commit a04cb61a). SkillShield performs automated 4-layer security analysis on AI skills and MCP servers.
Layer Breakdown
Behavioral Risk Signals
Security Findings1
| Severity | Finding | Layer | Location | |
|---|---|---|---|---|
| HIGH | Potential Data Exfiltration via Reading User Context File The skill instructs the host LLM to read a file named `.claude/product-marketing-context.md` if it exists. If the execution environment resolves this relative path to a user-specific or system-wide configuration directory (e.g., `~/.claude/`) rather than a file strictly within the skill's own sandboxed package, this could lead to unauthorized access and exfiltration of sensitive user context data. The content of such a file is unknown but could contain proprietary or personal information. The file is not provided as part of the skill's supporting files, suggesting it might be an external resource. 1. **Clarify File Scope**: Ensure the skill's file reading capabilities are strictly sandboxed to its own package directory. 2. **Bundle Context File**: If `product-marketing-context.md` is intended to be skill-specific, it should be included within the skill package and referenced with a path that clearly indicates it's internal (e.g., `skill_data/product-marketing-context.md`). 3. **User Consent/Input**: If the skill genuinely requires user-specific context, it should explicitly prompt the user for this information or utilize platform-provided, permission-controlled mechanisms for data access, rather than attempting to read files from potentially sensitive system locations. | Unknown | SKILL.md:8 |
Scan History
Embed Code
[](https://skillshield.io/report/bda6f54b23f1a661)
Powered by SkillShield