a2a-jsonrpc-transport

A2A JSON-RPC 2.0 Transport

Before writing code

Fetch live docs:

1. Fetch https://a2a-protocol.org/latest/specification/ for the transport layer section
2. Web-search site:github.com a2aproject A2A JSON-RPC transport for transport protocol details
3. Fetch https://www.jsonrpc.org/specification for the JSON-RPC 2.0 base specification
4. Fetch SDK docs for transport-level classes and middleware

Conceptual Architecture

Why JSON-RPC 2.0

A2A uses JSON-RPC 2.0 as its transport protocol because:

• Simple — Lightweight request/response format
• Standard — Well-established protocol with broad tooling support
• Language-agnostic — Works with any language that handles JSON over HTTP
• Extensible — Custom methods without changing the protocol

Request Format

Every A2A request is a JSON-RPC 2.0 request:

{
  "jsonrpc": "2.0",
  "method": "message/send",
  "id": "unique-request-id",
  "params": { ... }
}

• jsonrpc — Always "2.0"
• method — The A2A method name (e.g., message/send, tasks/get)
• id — Client-generated unique ID for matching responses
• params — Method-specific parameters (object)

Response Format

Success:

{
  "jsonrpc": "2.0",
  "id": "unique-request-id",
  "result": { ... }
}

Error:

{
  "jsonrpc": "2.0",
  "id": "unique-request-id",
  "error": {
    "code": -32600,
    "message": "Invalid Request",
    "data": { ... }
  }
}

HTTP Details

• Method: All A2A requests use HTTP POST
• URL: The agent's endpoint URL (from the Agent Card)
• Content-Type: application/json for regular requests
• Response Content-Type: application/json for message/send, text/event-stream for message/stream
• HTTP Status: Typically 200 OK — errors are in the JSON-RPC response body, not HTTP status codes

A2A Methods

| Method | Direction | Purpose | |--------|-----------|---------| | message/send | Client → Server | Send message, get synchronous response | | message/stream | Client → Server | Send message, get SSE stream | | tasks/get | Client → Server | Retrieve task by ID | | tasks/cancel | Client → Server | Cancel a task | | tasks/resubscribe | Client → Server | Re-subscribe to task's SSE stream | | tasks/pushNotificationConfig/set | Client → Server | Register push notification | | tasks/pushNotificationConfig/get | Client → Server | Get notification config | | tasks/pushNotificationConfig/list | Client → Server | List notification configs | | tasks/pushNotificationConfig/delete | Client → Server | Delete notification config | | agent/authenticatedExtendedCard | Client → Server | Get extended Agent Card |

Method Routing

The server must route incoming requests by the method field:

1. Parse the JSON body
2. Validate JSON-RPC 2.0 structure
3. Extract the method field
4. Route to the appropriate handler
5. Return the handler's result wrapped in a JSON-RPC response

Request IDs

• Client generates unique IDs for each request
• Server echoes the same ID in the response
• Use UUIDs or monotonically increasing integers
• IDs enable matching responses to requests in async scenarios

Idempotency

A2A doesn't mandate built-in idempotency, but best practice is:

• Use unique request IDs for deduplication
• Design handlers to be idempotent where possible
• For message/send, sending the same message twice should be handled gracefully

Best Practices

• Use the SDK's built-in JSON-RPC handling rather than implementing from scratch
• Validate the jsonrpc field is exactly "2.0"
• Return proper JSON-RPC errors for malformed requests (not HTTP error codes)
• Log request/response pairs with IDs for debugging
• Set appropriate HTTP timeouts (longer for message/stream)
• Use Content-Type: application/json for all non-streaming requests
• Handle unknown methods gracefully with -32601
• Don't rely on HTTP status codes for error handling — always check the JSON-RPC response body

Fetch the specification for any additional transport requirements, header conventions, and method parameter schemas before implementing.