cloudbase-agent-ts
Pass
Audited by Gen Agent Trust Hub on Mar 9, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external, untrusted data via the AG-UI protocol, creating a surface for indirect prompt injection attacks.
- Ingestion points: Untrusted data enters the agent context through the
messagesandtoolsfields in theRunAgentInputinterface (as seen inagui-protocol.md). - Boundary markers: The provided code examples (e.g., the
chatNodeimplementation inadapter-langgraph.md) interpolate user messages directly into the LLM invocation without using delimiters or instructions to ignore embedded commands. - Capability inventory: Agents built using this skill possess capabilities for network operations (communicating with LLM providers) and tool execution (both server-side and client-side as defined in
agui-protocol.md). - Sanitization: The documentation and examples do not include input validation or sanitization routines for the message content before it reaches the AI model.
Audit Metadata