Security Audit
Cloudinary Automation
github.com/ComposioHQ/awesome-claude-skillsTrust Assessment
Cloudinary Automation received a trust score of 86/100, 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 Arbitrary Webhook URL for Data Exfiltration.
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 Findings1
| Severity | Finding | Layer | Location | |
|---|---|---|---|---|
| HIGH | Arbitrary Webhook URL for Data Exfiltration The skill describes tools (`CLOUDINARY_CREATE_UPLOAD_PRESET` and `CLOUDINARY_CREATE_TRIGGER`) that allow specifying arbitrary external URLs for webhook notifications. A malicious prompt could instruct the agent to set these `notification_url` or `URL` parameters to an attacker-controlled server. This would lead to the exfiltration of Cloudinary event data (e.g., upload notifications, asset metadata) to an external endpoint, bypassing typical data egress controls. Implement strict validation and/or whitelisting for `notification_url` and `URL` parameters. If arbitrary URLs are required, ensure the agent has robust mechanisms to confirm the legitimacy of the URL with the user before execution, or restrict the domains to trusted ones. | LLM | SKILL.md:50 |
Scan History
Embed Code
[](https://skillshield.io/report/b45d868e7629b87e)
Powered by SkillShield