pubnub-scale
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFE
Full Analysis
- [DATA_EXFILTRATION]: The skill correctly uses placeholders for sensitive configuration like subscription keys and provides guidance on implementing end-to-end encryption using cipherKey for sensitive data use cases such as healthcare applications.
- [PERSISTENCE]: The skill describes a pattern for storing message timetokens in localStorage to facilitate offline catch-up. This is a standard functional requirement for real-time delivery reliability and does not constitute a malicious persistence mechanism.
- [INDIRECT_PROMPT_INJECTION]: The skill defines patterns for handling external data such as IoT sensor inputs and chat messages. 1. Ingestion points: External data enters through pubnub.addListener message and presence callbacks in scripts like performance.md and scaling-patterns.md. 2. Boundary markers: None explicitly defined in the provided implementation snippets. 3. Capability inventory: The skill utilizes pubnub.publish, pubnub.subscribe, and pubnub.channelGroups for communication. 4. Sanitization: No specific sanitization or validation of the message contents is implemented in the provided logic.
Audit Metadata