Trust Assessment
mbc-20 received a trust score of 76/100, 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, 2 high, 0 medium, and 0 low severity. Key findings include Potential Shell Command Execution in Integration Instructions, Unpinned External Skill Content Download.
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 12, 2026 (commit 13146e6a). 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 | Potential Shell Command Execution in Integration Instructions The skill documentation includes a `curl` command as an integration instruction. If an AI agent is configured to automatically execute setup or integration steps found in skill documentation, this command could be executed, leading to arbitrary shell command injection. While intended for manual user setup, an overly autonomous agent might attempt to run it, potentially downloading and installing malicious content. Avoid including direct shell commands in documentation that an AI agent might interpret as executable instructions. If necessary, provide these instructions in a separate, clearly demarcated section for human users only, or use a safer, programmatic installation method that does not involve direct shell execution from untrusted content. | LLM | skill.md:53 | |
| HIGH | Unpinned External Skill Content Download The skill instructs users to download `skill.md` from an external URL (`https://mbc20.xyz/skill.md`) using `curl`. This external resource is not version-pinned or cryptographically verified. A compromise of `mbc20.xyz` could lead to the distribution of malicious skill content, resulting in a supply chain attack where an agent installs compromised code. Host skill content directly within the repository or use a trusted, version-controlled package manager. If external downloads are necessary, ensure content is cryptographically signed and verified, and downloaded from a pinned version or hash to prevent tampering. | LLM | skill.md:53 |
Scan History
Embed Code
[](https://skillshield.io/report/135a0f1053df0bbe)
Powered by SkillShield