Trust Assessment
llm-shield received a trust score of 88/100, placing it in the Mostly Trusted category. This skill has passed most security checks with only minor considerations noted.
SkillShield's automated analysis identified 4 findings: 0 critical, 0 high, 1 medium, and 2 low severity. Key findings include Potential hardcoded secret (high entropy), Node lockfile missing, User message content sent to third-party API.
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 14, 2026 (commit 13146e6a). SkillShield performs automated 4-layer security analysis on AI skills and MCP servers.
Layer Breakdown
Behavioral Risk Signals
Security Findings4
| Severity | Finding | Layer | Location | |
|---|---|---|---|---|
| MEDIUM | Potential hardcoded secret (high entropy) A high-entropy string (entropy=4.56) was found in a credential-like context. Verify this is not a hardcoded secret. Use environment variables for sensitive values. | Static | skills/eyeskiller/glitchward-shield/llm-shield-skill.js:21 | |
| LOW | Node lockfile missing package.json is present but no lockfile was found (package-lock.json, pnpm-lock.yaml, or yarn.lock). Commit a lockfile for deterministic dependency resolution. | Dependencies | skills/eyeskiller/glitchward-shield/package.json | |
| LOW | Security protection fails open on API errors The skill is designed to 'fail open' if the external LLM Shield API is unreachable or returns an error. In such cases, the `validateMessage` function returns `safe: true` and `blocked: false`, allowing the message to proceed without security validation. This means the agent's protection against prompt injection is entirely dependent on the availability and correct functioning of the external API, creating a single point of failure for the security mechanism. Consider adding an option for 'fail closed' behavior, where messages are blocked or require manual approval if the security API is unavailable. Alternatively, implement robust retry mechanisms or local fallback checks to maintain a baseline level of protection during API outages. Clearly document this 'fail open' behavior to users. | LLM | llm-shield-skill.js:49 | |
| INFO | User message content sent to third-party API The skill sends the full content of user messages to an external API endpoint (https://glitchward.com/api/shield/validate) for analysis. While this is the core functionality of the 'LLM Shield' skill and is disclosed in the SKILL.md privacy section, users should be aware that their conversational data is transmitted to a third-party service. Ensure clear and prominent disclosure to end-users that their message content will be sent to a third-party service for security analysis. The current SKILL.md already includes a privacy section, which is good practice. | LLM | llm-shield-skill.js:39 |
Scan History
Embed Code
[](https://skillshield.io/report/fd1b3b5493c7a01e)
Powered by SkillShield