zoom-skills
Pass
Audited by Gen Agent Trust Hub on Mar 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscated code, or unauthorized data exfiltration mechanisms were identified in the analyzed repository files. The content is strictly instructional and developer-oriented.
- [EXTERNAL_DOWNLOADS]: The documentation references downloading SDK binaries and assets from trusted official sources, primarily
source.zoom.usandgithub.com/zoom. These downloads are recognized as safe developer dependencies according to the platform's trusted vendor list. - [PROMPT_INJECTION]: The skill definitions and triggers are specifically designed for routing Zoom-related development queries and do not contain patterns intended to bypass agent safety protocols or manipulate model behavior.
- [DATA_EXFILTRATION]: No evidence of unauthorized data transmission or hardcoded credentials was found. The repository correctly instructs developers to handle sensitive tokens and secrets using backend services and environment variables.
- [COMMAND_EXECUTION]: Code snippets for native SDKs (Windows and Linux) include necessary system-level operations such as the Windows message pump and GLib main loops. These are standard implementation details for the target platforms and do not represent malicious shell execution.
Audit Metadata