Langfuse MCP Server
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| LANGFUSE_BASEURL | No | Your Langfuse instance URL | https://us.cloud.langfuse.com |
| LANGFUSE_PUBLIC_KEY | Yes | Your Langfuse public key | |
| LANGFUSE_SECRET_KEY | Yes | Your Langfuse secret key |
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
Server capabilities have not been inspected yet.
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| list_projectsB | List configured Langfuse projects available to this MCP server. |
| project_overviewC | Get a summary of total cost, tokens, and traces for a project over a time window. |
| usage_by_modelC | Break down usage and cost by AI model over a time period. |
| usage_by_serviceC | Analyze usage and cost by service/feature tag over a time period. |
| top_expensive_tracesC | Find the most expensive traces by cost over a time period. |
| get_trace_detailC | Get detailed information about a specific trace including all observations. |
| get_projectsA | List available Langfuse projects (alias for list_projects). |
| get_metricsB | Query aggregated metrics (costs, tokens, counts) with flexible filtering and dimensions. |
| get_tracesC | Fetch traces with flexible filtering options. |
| get_observationsB | Get LLM generations/spans with details and filtering. |
| get_cost_analysisC | Specialized cost breakdowns by model, user, and daily trends. |
| get_daily_metricsC | Daily usage trends and patterns. |
| get_observation_detailC | Get detailed information about a specific observation by ID. |
| get_health_statusB | Get system health status and availability information. |
| list_modelsB | List all available AI models in the Langfuse project. |
| get_model_detailB | Get detailed information about a specific AI model. |
| list_promptsC | List all prompt templates in the Langfuse project. |
| get_prompt_detailC | Get detailed information about a specific prompt template. |
| list_datasetsB | List all datasets in the project with pagination support. |
| get_datasetC | Get detailed information about a specific dataset by name. |
| list_dataset_itemsC | List items in datasets with filtering and pagination. |
| get_dataset_itemC | Get detailed information about a specific dataset item. |
| list_commentsC | List comments with filtering options for objects and users. |
| get_commentC | Get detailed information about a specific comment. |
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 24 tools
Most tools have distinct purposes targeting specific resources (e.g., get_comment vs. get_metrics), but some overlap exists between get_observations and get_trace_detail (since traces include observations) and between get_daily_metrics and project_overview (both provide aggregated metrics). Descriptions help clarify, but agents might occasionally misselect between these related tools.
Tool names follow a highly consistent verb_noun pattern throughout, with 'get_' for retrieving specific items, 'list_' for listing collections, and descriptive nouns (e.g., get_trace_detail, list_datasets). There are no deviations in naming conventions, making the set predictable and readable.
With 24 tools, the count is borderline high for a single server, potentially overwhelming for agents. While the Langfuse domain is broad (covering traces, metrics, datasets, etc.), the tool set feels heavy and could benefit from consolidation or better scoping to reduce cognitive load.
The tool surface provides comprehensive coverage for the Langfuse domain, including CRUD-like operations (get/list for traces, observations, datasets, prompts, models, comments), detailed metrics (cost, usage, health), and specialized analyses (top_expensive_traces, usage_by_model). No obvious gaps exist; agents can perform full lifecycle and analytical workflows.