The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were identified in the skill files. The content consists of technical documentation, API references, and code examples for implementing the AG-UI protocol. No sensitive credentials, unauthorized network exfiltration, or persistence mechanisms were found.
[INDIRECT_PROMPT_INJECTION]: The protocol specification defines ingestion points for untrusted data, specifically via the messages, state, and context fields in the RunAgentInput object (documented in references/protocol-spec.md). While this represents a surface for indirect prompt injection common to agent protocols, the documentation correctly defines a structured, typed event-based framework for managing these interactions. Capability inventory includes LLM-based response generation (generateResponse in references/building-agents.md) and tool execution, which are standard for the intended use-case. No sanitization logic is explicitly provided in the examples, which is expected for protocol-level documentation.

copilotkit-agui