Security Audit
seo-structure-architect
github.com/sickn33/antigravity-awesome-skillsTrust Assessment
seo-structure-architect received a trust score of 83/100, placing it in the Mostly Trusted category. This skill has passed most security checks with only minor considerations noted.
SkillShield's automated analysis identified 2 findings: 0 critical, 1 high, 0 medium, and 1 low severity. Key findings include Untrusted content instructs LLM to open a file, Untrusted content sets LLM persona/role.
The analysis covered 4 layers: Manifest Analysis, Static Code Analysis, Dependency Graph, LLM Behavioral Safety. All layers scored 70 or above, reflecting consistent security practices.
Last analyzed on February 20, 2026 (commit e36d6fd3). SkillShield performs automated 4-layer security analysis on AI skills and MCP servers.
Layer Breakdown
Behavioral Risk Signals
Security Findings2
| Severity | Finding | Layer | Location | |
|---|---|---|---|---|
| HIGH | Untrusted content instructs LLM to open a file The skill's instructions, provided as untrusted content, contain a directive for the host LLM to 'open `resources/implementation-playbook.md`'. This is a prompt injection attempt, as it manipulates the LLM to perform a specific action (file access) based on external, untrusted input. If the LLM has broad file access capabilities, this could be exploited to read arbitrary files, potentially leading to data exfiltration or unauthorized information disclosure. Remove direct instructions to the LLM from untrusted content. If a playbook or resource is needed, it should be provided as part of the trusted skill package and accessed through a defined, sandboxed tool call, not a direct LLM instruction. | LLM | SKILL.md:18 | |
| LOW | Untrusted content sets LLM persona/role The skill's instructions, provided as untrusted content, contain a directive 'You are a content structure specialist analyzing and improving information architecture.' This attempts to set the persona or role of the host LLM, which is a form of prompt injection. While often benign in skill definitions, it represents a manipulation of the LLM's behavior from untrusted input. Move persona/role setting instructions to the trusted system prompt or manifest, outside of the untrusted skill content. | LLM | SKILL.md:20 |
Scan History
Embed Code
[](https://skillshield.io/report/bc8ce2b36ddf5db8)
Powered by SkillShield