managedcode-orleans-signalr
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious behavior or obfuscation detected. The skill references official packages and source repositories belonging to the author 'managedcode' and uses the trusted '@microsoft/signalr' library for client-side communication.
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates distributed message delivery, creating a surface where the agent processes external data from SignalR clients.
- Ingestion points: Data enters through the
UpdatesHubendpoint and client-side message receivers inSKILL.md. - Boundary markers: The documentation includes clear instructions to derive authorization and identity from server-side authenticated state rather than client-supplied parameters.
- Capability inventory: The skill enables network-based publishing to specific users and groups via
IHubContext. - Sanitization: The instructions explicitly warn developers to validate identities and implement idempotent handlers to prevent unauthorized data routing or replay attacks.
Audit Metadata