nxtlinq-mcp-proxy

# nxtlinq MCP Proxy Package an existing MCP server, deploy it on a private Docker network, and check nxtlinq tool permissions before forwarding calls. The original MCP implements the tools. Desktop Setup manages identity and tool permissions. ## Requirements - uv on macOS or Linux; the guide uses uv to provide Python 3.11. - Docker with Compose and Buildx, running on the deployment host. - A dedicated nxtlinq thirdParty service API Key and Secret. - An Agent supporting Streamable HTTP MCP with an Authorization header. Automatic Codex configuration is an optional integration. ## Install Confirm `uv --version` works and that `docker info`, `docker compose version`, and `docker buildx version` succeed. The source distribution includes step-by-step environment installation guides at `docs/en/environment.md` and `docs/zh-TW/environment.md`; skip components that already pass these checks. In a new working directory, create a Python virtual environment with uv, then install the Proxy and its dependencies. uv downloads Python 3.11 if needed. For an existing installation, activate its environment and skip environment creation. ```sh mkdir nxtlinq-tools cd nxtlinq-tools uv venv --python 3.11 .venv source .venv/bin/activate ``` Install from PyPI: ```sh uv pip install 'nxtlinq-mcp-proxy[desktop]==0.1.0' ``` uv downloads the Proxy and its dependencies from PyPI. `[desktop]` includes Setup dependencies. Download: [Proxy 0.1.0](https://pypi.org/project/nxtlinq-mcp-proxy/0.1.0/#files). ## Deploy ```sh nxtlinq-mcp pack my-mcp-bundle nxtlinq-mcp init my-tools --bundle my-mcp-bundle cd my-tools nxtlinq-mcp setup ``` Choose the packaging method for your MCP. During initialization, select `other` for manual MCP configuration or `codex` for the optional automatic integration. Enter the service environment and credentials when prompted, review the discovered tools, and confirm registration. Complete Setup and enable the tools you need, then connect your Agent using the ch…