Metadata-Version: 2.4
Name: lmrs-client
Version: 0.1.34
Summary: Client library for the Local Model Runtime Service (LMRS)
Author-email: Vasiliy Zdanovskiy <vasilyvz@gmail.com>
License: MIT
Requires-Python: >=3.12
Description-Content-Type: text/markdown
Requires-Dist: mcp-proxy-adapter>=8.11.10

# lmrs-client

Client library and CLI for the Local Model Runtime Service (LMRS).

Each client method mirrors one public LMRS command. Transport, TLS and error
mapping are handled by the client; callers never build URLs or JSON-RPC payloads.

## Install

    pip install lmrs-client

## Usage

    import asyncio
    from lmrs_client import LmrsClient

    async def main():
        client = LmrsClient(host="lmrs", port=8012, protocol="https", ca="/etc/lmrs/ca.crt")
        info = await client.current_model()
        print(info["model"], info["context_window"])

    asyncio.run(main())

`current_model()` reports the model the running vLLM serves and its context window.

## Command line

    lmrs-client --help

Each client method is a subcommand; the exit status follows the command verdict.

## Timeouts

The client timeout is independent of the server's `timeouts` config section.
Resolution: constructor `timeout`, then the environment variable
`MCP_PROXY_ADAPTER_HTTP_TIMEOUT`, then 30.0 s; the CLI uses `--timeout` /
`LMRS_LIVE_TIMEOUT`. The server's vLLM call timeout (`timeouts.runtime_seconds`,
default 180 s) may be longer than the client default, so a long chat can end on
the client side first; raise the client timeout for such calls. `info` reports
the server's effective values in `runtime_summary.timeouts`.
