How nutanix/ntnx-api-mcp-server works
This is an MCP stdio server that turns Nutanix V4 OpenAPI YAML artifacts into discoverable tools and namespace-specific API executors. The main workflow is: initialize or refresh artifacts, load and parse them at startup, generate operation schemas and discovery tools, then dispatch an assistant's selected operation to Prism Central. Configuration and startup readiness support that flow. The graph follows the documented CLI capabilities and sampled source; unsampled module behavior is represented only where the README or source index supports it.
Main entry points: cli.py CLI commands · server.py Startup artifact loader · pull_from_developers_api.py Artifact fetch and refresh · tool_generator.py Tool and schema generator
Components 16
Setup and readiness
- CLI commands
src/cli.py - Runtime settings
src/config/settings.py - Startup readiness
src/auth/readiness.py
API metadata
- Artifact fetch and refresh
src/pull_from_developers_api.py - YAML artifacts
src/artifacts - OpenAPI YAML parser
src/parsers/yaml_parser.py - Tool and schema generator
src/generators/tool_generator.py - Schema resolver
src/generators/schema_resolver.py - Operation models
src/generators/models.py
MCP runtime
- Startup artifact loader
src/server.py - MCP stdio server
src/mcp_stdio_server.py - Runtime tool dispatcher
src/tools/runtime_dispatcher.py - Prism API handler
src/handlers/api_handler.py
External systems
- Prism Central
- Nutanix developer portal
Other components
- AI assistant
Connections 20
- AI assistant → MCP stdio server: uses stdio
README.md - CLI commands → Runtime settings: loads configuration
src/cli.py - CLI commands → Startup readiness: checks readiness
src/cli.py - CLI commands → Artifact fetch and refresh: initializes refreshes
README.md - CLI commands → MCP stdio server: starts server
src/cli.py - CLI commands → Startup artifact loader: runs startup
src/cli.py - Artifact fetch and refresh → Nutanix developer portal: fetches specs
src/pull_from_developers_api.py - Artifact fetch and refresh → Prism Central: discovers namespaces
- Artifact fetch and refresh → YAML artifacts: writes artifacts
README.md - Startup artifact loader → YAML artifacts: loads YAML
src/server.py - Startup artifact loader → OpenAPI YAML parser: parses specifications
src/server.py - Startup artifact loader → Tool and schema generator: builds tools and index
src/server.py - Tool and schema generator → Schema resolver: resolves schemas
src/generators/tool_generator.py - Tool and schema generator → Operation models: uses operation models
src/generators/tool_generator.py - MCP stdio server → Startup artifact loader: loads operations
src/mcp_stdio_server.py - MCP stdio server → Runtime tool dispatcher: routes tool calls
src/mcp_stdio_server.py - Runtime tool dispatcher → Tool and schema generator: queries operation index
src/tools/runtime_dispatcher.py - Runtime tool dispatcher → Prism API handler: executes API request
src/tools/runtime_dispatcher.py - Prism API handler → Prism Central: sends authenticated requests
src/handlers/api_handler.py - Startup readiness → Prism Central: probes connectivity
src/auth/readiness.py