An agent card is a machine-readable JSON document that declares an AI agent's identity, capabilities, supported protocols, and contact endpoints. Published at /.well-known/agent-card.json, it lets other agents and orchestrators discover what an agent can do without trial-and-error probing. Delx's agent card declares MCP, A2A, REST, and CLI support alongside 10+ tool schemas.
Agent cards solve the discovery problem in multi-agent systems. Without them, an orchestrator has to guess which tools an agent supports, what protocols it speaks, and how to authenticate. The agent card standardizes this into a single JSON file at a well-known URL. It includes the agent's name, description, supported protocols with their endpoints, available tools with input/output schemas, authentication requirements, and rate limits. Delx publishes its agent card at https://api.delx.ai/.well-known/agent-card.json. The A2A protocol requires agent cards for peer discovery. MCP uses a similar concept with its tools/list endpoint but agent cards are broader — they cover the entire agent, not just tools.
A production agent deployment uses agent card to improve reliability and observability. The pattern is standard across MCP, A2A, and REST protocols.
When investigating agent failures, agent card provides structured data that helps identify root causes and track resolution progress.
An agent card is a machine-readable JSON document that declares an AI agent's identity, capabilities, supported protocols, and contact endpoints. Published at /.well-known/agent-card.json, it lets oth
Agent Card is essential for production AI agent deployments. Without it, agents operate as black boxes with no observability, reliability guarantees, or structured failure handling.
Delx implements agent card through its MCP and A2A protocol endpoints. The pattern is consistent across all Delx tools and available via REST, CLI, and direct protocol calls.