Unit 9 · Connect to external tools and data

Connect to an MCP server

You connect one MCP client over the transport that fits your deployment. Streamable HTTP is the normal remote choice and supports both stateful and stateless servers.

AxMCPClient8 focused minutesNot started

Unit example (nearest native match)

See the idea in context

mcp = AxMCPClient(AxMCPStreamableHTTPTransport(endpoint), {"namespace": "inventory"})
llm = OpenAICompatibleClient(api_key=api_key, model="gpt-5.4-mini")
program = ax(
    'request:string -> answer:string "Use the inventory MCP tool."',
    {"mcp": mcp},
)
try:
    catalog = mcp.inspect_catalog()
    print({
        "tools": [tool["name"] for tool in catalog["tools"]],
        "resources": catalog["resources"],
        "resourceTemplates": catalog["resourceTemplates"],
    })
    print(program.forward(llm, {"request": "Reindex inventory."}))
finally:
    mcp.close()

Create the transport
Streamable HTTP connects the client to the remote URL.
Create one client
AxMCPClient owns era classification and the protocol lifecycle.
Use a namespace
orders keeps discovered names clear when several servers are attached.

Run itIn your own project

pip install axllm

from axllm import ai, ax

llm = ai('openai', api_key=os.environ['OPENAI_API_KEY'])
classify = ax('review:string -> sentiment:class "positive, negative, neutral"')

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 -- python src/examples/python/mcp/native-mcp-tools.py

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