web-realtime-websockets
Pass
Audited by Gen Agent Trust Hub on Oct 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill implements communication patterns with external servers using WebSockets, which constitutes an ingestion point for untrusted data that could potentially contain malicious instructions.
- Ingestion points:
socket.onmessagehandlers inSKILL.md,examples/core.md, andexamples/presence.md. - Boundary markers: The skill promotes the use of discriminated unions (Pattern 5) to provide structural validation of incoming messages, which acts as a structural boundary.
- Capability inventory: The skill is focused on communication and state management; it does not implement dangerous capabilities such as arbitrary command execution, file system writes, or subprocess spawning.
- Sanitization: Examples utilize
JSON.parse()for data extraction and implement structural checks to validate message types before processing content. - [EXTERNAL_DOWNLOADS]: The skill metadata references a JSON schema located in the official GitHub repository of the vendor (
agents-inc). This is a legitimate reference to vendor-owned infrastructure for configuration validation.
Audit Metadata