jazz-ui-development

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEDATA_EXFILTRATIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • Data Exposure & Exfiltration (LOW): The skill configures network connections to wss://cloud.jazz.tools. While this is the primary intended function for synchronization, it constitutes a network operation to a non-whitelisted domain. All sensitive placeholders like ${apiKey} are correctly identified as user-provided variables, and no actual credentials are exposed.
  • Indirect Prompt Injection (LOW): The skill establishes an attack surface for indirect prompt injection by facilitating the ingestion of data from remote peers and user inputs (via useCoState, useAccount, and CoValueSchema.subscribe).
  • Ingestion points: Data entering via useCoState and useAccount in references/react.md, references/svelte.md, and references/vanilla.md.
  • Boundary markers: Absent; the documentation focuses on rendering and state management without explicit instructions for sanitizing or delimiting untrusted data content.
  • Capability inventory: The described tools primarily perform UI rendering and state management; however, they enable complex data fetching and property updates.
  • Sanitization: Not explicitly detailed in the instructional content.
  • External Downloads (SAFE): The documentation references external packages (e.g., react-native-passkey, svelte-clerk) and documentation URLs on jazz.tools. These are standard for the described framework and do not involve automated script execution or untrusted downloads within the skill's own operational context.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:41 PM