Unit 9 · Connect to external tools and data

Classify stateful and stateless MCP

Automatic era classification keeps one client compatible with stateful 2025-11-25 and stateless 2026-07-28 servers. Modern discovery returns supported versions, capabilities, instructions, cache policy, and server identity.

discover()9 focused minutesNot started

Unit example (nearest native match)

See the idea in context

auto transport = std::make_shared<axllm::AxMCPStreamableHTTPTransport>(endpoint);
auto mcp = std::make_shared<axllm::AxMCPClient>(transport, axllm::object({{"namespace", "inventory"}}));
axllm::AxExecutionContext context({mcp});
auto program = axllm::ax("request:string -> answer:string");
context.attach(program);
axllm::OpenAICompatibleClient llm(axllm::object({
  {"api_key", key}, {"model", "gpt-5.4-mini"},
}));
auto catalog = mcp->inspect_catalog();
std::cout << "MCP catalog: " << axllm::Core::iter(catalog.tools).size()
          << " tools, " << axllm::Core::iter(catalog.resources).size()
          << " resources\n";
std::cout << axllm::stringify(program.forward(
  llm, axllm::object({{"request", "Reindex inventory."}})
)) << "\n";
mcp->close();

Keep automatic classification
Pin an era only when deployment policy already knows the endpoint contract.
Read modern discovery
server/discover is stateless and is available only after the endpoint classifies as modern.

Run itIn your own project

cmake_minimum_required(VERSION 3.20)
project(ax_quick_start LANGUAGES CXX)
set(CMAKE_CXX_STANDARD 17)

include(FetchContent)
FetchContent_Declare(axllm GIT_REPOSITORY https://github.com/ax-llm/ax GIT_TAG main SOURCE_SUBDIR packages/cpp)
FetchContent_MakeAvailable(axllm)

add_executable(quick_start quick_start.cpp)
target_link_libraries(quick_start PRIVATE axllm::axllm)

#include <axllm/axllm.hpp>
#include <cstdlib>
#include <iostream>

int main() {
  auto llm = axllm::ai("openai", axllm::object({{"apiKey", std::getenv("OPENAI_API_KEY")}}));
  auto classify = axllm::ax("review:string -> sentiment:class \"positive, negative, neutral\"");
  auto result = classify.forward(*llm, axllm::object({
    {"review", "Useful and boring in the best way."}
  }));

  auto sentiment = axllm::Core::get(result, "sentiment");
  std::cout << "sentiment: " << std::get<std::string>(sentiment.data) << "\n";
}

Set OPENAI_APIKEY in your environment before running provider-backed code.

In the ax repo

From a clone of the ax repo:

npm run example -- cpp src/examples/cpp/mcp/native_mcp_tools.cpp

Active practice

Show that you can use it

Answer 2 in a row to learn this · attempt 1

Keep exploring

See the idea in context

Show that you can use it

Source-backed follow-up