MLflow MCP Server
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| MLFLOW_TRACKING_URI | Yes | The MLflow tracking URI to connect to your MLflow instance |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| get_experimentC | Get experiment details |
| get_experiment_by_nameC | Get experiment details by name |
| search_experimentsC | List all experiments |
| get_runC | Get a run by ID |
| get_experiment_runsC | Get all runs for an experiment |
| get_registered_modelsD | – |
| get_model_versionsD | – |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 7 tools
Most tools have distinct purposes targeting different MLflow entities (experiments, runs, models), but 'get_experiment' and 'get_experiment_by_name' could cause confusion as they serve the same ultimate purpose through different lookup methods. The other tools clearly target different resources.
All tools follow a consistent verb_noun pattern with 'get_' or 'search_' prefixes, using snake_case throughout. The naming is predictable and follows a clear convention across all seven tools.
Seven tools is well-scoped for an MLflow server, covering key entities like experiments, runs, and models. The count feels appropriate—neither too sparse nor bloated—for the apparent domain scope.
The toolset provides good read/search coverage for experiments, runs, and models, but lacks any write operations (create, update, delete) or lifecycle management tools. This is a notable gap that will limit agent workflows to read-only scenarios.