Streaming End-to-End MCP Tool Use with an AI Model

See more AI Examples

Demonstrates Ai.UseMcp in a streaming Model Context Protocol (MCP) workflow. This is the same idea as the non-streaming MCP example, but the answer is streamed token-by-token and the automatic MCP tool call is driven by the application's event loop rather than happening invisibly inside Ask.

A tool-using conversation spans more than one streamed turn. In the first turn the model streams a js_function_call event; the example hands that event's JSON to StreamingJsToolCall, which routes MCP tools to the server and appends the result to the transcript, and the turn ends with null_terminator. The example then calls Ask again (with no new input, since the tool result is already in the transcript) so the model can continue and stream its final answer. The loop ends when a streamed turn finishes with no tool call outstanding.

Tip: Code to parse the returned JSON can be generated with Chilkat's online tool at https://tools.chilkat.io/jsonParse.

llSuccess = .F.

//  Streaming, end-to-end MCP example.  Same idea as the non-streaming MCP example, but the answer is
//  streamed token-by-token, and the automatic MCP tool call is driven by the application's event loop
//  instead of happening invisibly inside Ask.
//  
//  A tool-using conversation spans more than one streamed turn:
//  
//    turn 1:  the model streams a "js_function_call" event (it wants a tool).  We hand that event's
//             JSON to StreamingJsToolCall, which routes MCP tools to the server (Mcp.CallTool) and
//             appends the result to the transcript.  The turn then ends with "null_terminator".
//    turn 2:  we call Ask again (no new input -- the tool result is already in the transcript) so the
//             model can continue.  It streams the final answer and ends with "null_terminator", with
//             no pending tool call.
//  
//  The loop ends when a streamed turn finishes with no tool call outstanding.
//  
//  MCP server: DeepWiki -- a free, public, no-authentication Streamable HTTP server that answers
//  questions about public GitHub repositories.  Swap in any URL; set mcp.AuthToken before Connect if
//  it needs a bearer token.

//  1) Connect to the MCP server.
loMcp = createobject("CkMcp")
//  mcp.AuthToken = "...";   // <-- only if the server requires a bearer token
llSuccess = loMcp.Connect("https://mcp.deepwiki.com/mcp")
if (llSuccess = .F.) then
    ? loMcp.LastErrorText
    release loMcp
    return
endif

? "Connected to MCP server: " + loMcp.ServerName + " (version " + loMcp.ServerVersion + ")"

//  2) Set up the AI conversation and register the MCP tools.
loAi = createobject("CkAi")
loAi.Provider = "anthropic"
//  The API key should come from a secure source rather than being hard-coded.
loAi.ApiKey = "AI_PROVIDER_API_KEY"
loAi.Model = "claude-sonnet-5"

//  Automatic tool use requires a conversation (not a stateless query).
llSuccess = loAi.NewConvo("test_conversation","You are a helpful assistant.  Use the available tools when they can help answer the question.","")
if (llSuccess = .F.) then
    ? loAi.LastErrorText
    release loMcp
    release loAi
    return
endif

//  Namespace the server's tools with "deepwiki" -- e.g. "deepwiki_ask_question".  The Mcp object must
//  stay connected and in scope for the whole conversation.
llSuccess = loAi.UseMcp(loMcp,"deepwiki")
if (llSuccess = .F.) then
    ? loAi.LastErrorText
    release loMcp
    release loAi
    return
endif

//  Multiple MCP servers can be used in the same conversation.  Connect each server and call UseMcp
//  once per server, giving each a distinct prefix so their tool names cannot collide.  For example:
//      success = ai.UseMcp(mcpWeather,"weather");
//      success = ai.UseMcp(mcpGitHub,"github");

llSuccess = loAi.InputAddText('Use the available tools to look up the GitHub repository "modelcontextprotocol/modelcontextprotocol", then give me a brief, sourced summary of which transports the Model Context Protocol defines.')
if (llSuccess = .F.) then
    ? loAi.LastErrorText
    release loMcp
    release loAi
    return
endif

//  3) Start streaming.
loAi.Streaming = .T.
llSuccess = loAi.Ask("text")
if (llSuccess = .F.) then
    ? loAi.LastErrorText
    release loMcp
    release loAi
    return
endif

//  4) Drive the streaming event loop.
loSbEventName = createobject("CkStringBuilder")
loSbDelta = createobject("CkStringBuilder")
loSbFullResponse = createobject("CkStringBuilder")
llBConversationDone = .F.
llBToolCallPending = .F.
llBCaseSensitive = .T.
llBAbort = .F.
lnLoopGuard = 0
lnResult = 0

do while (llBConversationDone <> .T.) and (lnLoopGuard < 8000)
    lnLoopGuard = lnLoopGuard + 1
    lnResult = loAi.PollAi(llBAbort)
    if (lnResult < 0) then
        ? loAi.LastErrorText
        release loMcp
        release loAi
        release loSbEventName
        release loSbDelta
        release loSbFullResponse
        return
    endif

    if (lnResult = 0) then
        //  No event ready yet.
        loAi.SleepMs(100)
    else
        llSuccess = loAi.NextAiEvent(5000,loSbEventName,loSbDelta)
        if (llSuccess = .F.) then
            ? loAi.LastErrorText
            release loMcp
            release loAi
            release loSbEventName
            release loSbDelta
            release loSbFullResponse
            return
        endif

        if (loSbEventName.ContentsEqual("js_function_call",llBCaseSensitive) = .T.) then
            //  (a) The model is requesting a tool call.  sbDelta holds the function-call JSON.
            //  StreamingJsToolCall runs the tool -- routing MCP tools to the server -- and appends the
            //  result to the transcript.  It must be called before this turn's null_terminator.
            ? "[tool call requested]"
            ? loSbDelta.GetAsString()
            llSuccess = loAi.StreamingJsToolCall(loSbDelta)
            if (llSuccess = .F.) then
                ? loAi.LastErrorText
                release loMcp
                release loAi
                release loSbEventName
                release loSbDelta
                release loSbFullResponse
                return
            endif

            llBToolCallPending = .T.
        else
            if (loSbEventName.ContentsEqual("null_terminator",llBCaseSensitive) = .T.) then
                //  (c) This streamed turn has finished.
                if (llBToolCallPending = .T.) then
                    //  Continue the conversation so the model can use the tool result.  No new
                    //  InputAddText -- the tool result is already in the transcript.
                    llBToolCallPending = .F.
                    llSuccess = loAi.Ask("text")
                    if (llSuccess = .F.) then
                        ? loAi.LastErrorText
                        release loMcp
                        release loAi
                        release loSbEventName
                        release loSbDelta
                        release loSbFullResponse
                        return
                    endif

                else
                    llBConversationDone = .T.
                endif

            else
                //  (b) An "empty" event reports nothing this poll.  (d) Any other event is a normal
                //  streamed text delta, which we display and accumulate.
                if (loSbEventName.ContentsEqual("empty",llBCaseSensitive) <> .T.) then
                    ? "Event: " + loSbEventName.GetAsString() + "  Delta: " + loSbDelta.GetAsString()
                    loSbFullResponse.AppendSb(loSbDelta)
                endif

            endif

        endif

    endif

enddo

? "Final streamed response:"
? loSbFullResponse.GetAsString()
? "----"

//  (Optional) The full transcript, including the tool call and its result.
loConvoJson = createobject("CkJsonObject")
loConvoJson.EmitCompact = .F.
llSuccess = loAi.ExportConvo("test_conversation",loConvoJson)
if (llSuccess = .F.) then
    ? loAi.LastErrorText
    release loMcp
    release loAi
    release loSbEventName
    release loSbDelta
    release loSbFullResponse
    release loConvoJson
    return
endif

? "Full Conversation:"
? loConvoJson.Emit()

//  5) Close the MCP session.
llSuccess = loMcp.Close()
if (llSuccess = .F.) then
    ? loMcp.LastErrorText
    release loMcp
    release loAi
    release loSbEventName
    release loSbDelta
    release loSbFullResponse
    release loConvoJson
    return
endif



release loMcp
release loAi
release loSbEventName
release loSbDelta
release loSbFullResponse
release loConvoJson