Trust Assessment
box-automation received a trust score of 73/100, placing it in the Caution category. This skill has some security considerations that users should review before deployment.
SkillShield's automated analysis identified 2 findings: 0 critical, 2 high, 0 medium, and 0 low severity. Key findings include Excessive Permissions: Broad Box Account Control, Data Exfiltration Risk: File Download Capability.
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 17, 2026 (commit 99e2a295). 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 | Excessive Permissions: Broad Box Account Control The skill grants the AI agent extensive control over the user's Box account, including capabilities to upload, download, delete, and permanently remove files and folders, as well as manage sharing and collaborations. While these are core functions of a Box automation skill, they represent a significant security risk if the agent is compromised or misused. A malicious agent could delete critical data, exfiltrate sensitive information, or grant unauthorized access to Box content. Implement strict access controls and user consent mechanisms for sensitive operations. Ensure the agent's execution environment is highly secure and isolated. Consider breaking down the skill into smaller, more granular skills with limited scopes if full Box control is not always necessary. Require explicit user confirmation for destructive actions like permanent deletion or broad sharing changes. | LLM | SKILL.md:105 | |
| HIGH | Data Exfiltration Risk: File Download Capability The skill explicitly provides the `BOX_DOWNLOAD_FILE` tool, allowing the AI agent to download files from the user's Box account. This capability, while essential for Box automation, poses a direct data exfiltration risk. If a malicious actor gains control of the agent, they could use this tool to download sensitive documents and transfer them outside the user's control. Additionally, `BOX_LIST_ITEMS_IN_FOLDER` and `BOX_SEARCH_FOR_CONTENT` can be used to discover sensitive data before exfiltration. Implement robust logging and auditing for all file download operations. Require explicit user approval or multi-factor authentication for downloading highly sensitive files. Restrict the agent's access to specific folders or file types where possible. Monitor for unusual download patterns or large data transfers initiated by the agent. | LLM | SKILL.md:50 |
Scan History
Embed Code
[](https://skillshield.io/report/6c2fe612c9519143)
Powered by SkillShield