base44-platform-actors

Pass

Audited by Gen Agent Trust Hub on Nov 3, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill provides legitimate developer guidance for building WebSocket-based realtime applications on the Base44 platform, including architectural patterns and contract definitions.
  • [INDIRECT_PROMPT_INJECTION]: The skill identifies a surface where actors process untrusted data from external clients, but provides appropriate defensive guidance.
  • Ingestion points: The handleMessage function in references/web.md and references/mobile.md processes arbitrary JSON payloads from WebSocket connections.
  • Boundary markers: The documentation explicitly warns that "the wire payload is attacker-controlled" and mandates runtime validation of all messages.
  • Capability inventory: Actor scripts have access to persistent storage (this.storage) and administrative service roles (this.client.asServiceRole) for interacting with other platform entities and functions.
  • Sanitization: Code examples demonstrate practical sanitization techniques, such as input length capping, type enforcement, and the use of explicit public projections for broadcasts to prevent leaking internal server state or secrets.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 3, 2026, 02:21 PM