Langfuse API MCP
A stateless MCP server providing AI agents with access to the full Langfuse public API, supporting both Langfuse Cloud and self-hosted instances. It enables agents to investigate traces, observations, sessions, errors, latency, and cost, as well as read prompts, datasets, experiments, scores, metrics, models, comments, and annotation queues. It also supports write operations like promoting prompt labels, creating scores, or editing datasets when explicitly enabled. The server is designed for corporate networks that re-sign TLS traffic with their own certificate authority.
Package and access
Local package with source-attributed installation metadata. Koskamo does not execute packages during review.
What you can do with it
- Investigate traces, observations, sessions, errors, latency, and cost
- Read prompts, datasets, experiments, scores, metrics, models, comments and annotation queues
- Promote a prompt label, create a score or edit a dataset
Source and maintenance
- Maintenance checked
Packages and installation
Source-provided package data. Koskamo has not executed or live-probed this package.
ghcr.io/rodrigorjsf/langfuse-api-mcp:0.2.0
{
"transport": "stdio",
"identifier": "ghcr.io/rodrigorjsf/langfuse-api-mcp:0.2.0",
"runtimeHint": "docker",
"registryType": "oci",
"environmentVariables": [
{
"name": "LANGFUSE_PUBLIC_KEY",
"isRequired": true,
"description": "The project's public key (pk-lf-...)."
},
{
"name": "LANGFUSE_SECRET_KEY",
"isRequired": true,
"description": "The project's secret key (sk-lf-...)."
},
{
"name": "LANGFUSE_BASE_URL",
"isRequired": true,
"description": "Your Langfuse host, e.g. https://cloud.langfuse.com; https only, http only for localhost."
}
]
}Tools
Reported by the attributed source; not independently observed by Koskamo.
Connect from your client
Each snippet configures the same source-provided package; replace placeholders before use.
Cursor
{
"mcpServers": {
"langfuse-api-mcp": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e",
"LANGFUSE_PUBLIC_KEY=<YOUR_LANGFUSE_PUBLIC_KEY>",
"-e",
"LANGFUSE_SECRET_KEY=<YOUR_LANGFUSE_SECRET_KEY>",
"-e",
"LANGFUSE_BASE_URL=<YOUR_LANGFUSE_BASE_URL>",
"ghcr.io/rodrigorjsf/langfuse-api-mcp:0.2.0"
],
"env": {
"LANGFUSE_PUBLIC_KEY": "<YOUR_LANGFUSE_PUBLIC_KEY>",
"LANGFUSE_SECRET_KEY": "<YOUR_LANGFUSE_SECRET_KEY>",
"LANGFUSE_BASE_URL": "<YOUR_LANGFUSE_BASE_URL>"
}
}
}
}Cline
{
"mcpServers": {
"langfuse-api-mcp": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e",
"LANGFUSE_PUBLIC_KEY=<YOUR_LANGFUSE_PUBLIC_KEY>",
"-e",
"LANGFUSE_SECRET_KEY=<YOUR_LANGFUSE_SECRET_KEY>",
"-e",
"LANGFUSE_BASE_URL=<YOUR_LANGFUSE_BASE_URL>",
"ghcr.io/rodrigorjsf/langfuse-api-mcp:0.2.0"
],
"env": {
"LANGFUSE_PUBLIC_KEY": "<YOUR_LANGFUSE_PUBLIC_KEY>",
"LANGFUSE_SECRET_KEY": "<YOUR_LANGFUSE_SECRET_KEY>",
"LANGFUSE_BASE_URL": "<YOUR_LANGFUSE_BASE_URL>"
}
}
}
}VS Code
{
"servers": {
"langfuse-api-mcp": {
"type": "stdio",
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e",
"LANGFUSE_PUBLIC_KEY=<YOUR_LANGFUSE_PUBLIC_KEY>",
"-e",
"LANGFUSE_SECRET_KEY=<YOUR_LANGFUSE_SECRET_KEY>",
"-e",
"LANGFUSE_BASE_URL=<YOUR_LANGFUSE_BASE_URL>",
"ghcr.io/rodrigorjsf/langfuse-api-mcp:0.2.0"
],
"env": {
"LANGFUSE_PUBLIC_KEY": "<YOUR_LANGFUSE_PUBLIC_KEY>",
"LANGFUSE_SECRET_KEY": "<YOUR_LANGFUSE_SECRET_KEY>",
"LANGFUSE_BASE_URL": "<YOUR_LANGFUSE_BASE_URL>"
}
}
}
}Claude Desktop
{
"mcpServers": {
"langfuse-api-mcp": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e",
"LANGFUSE_PUBLIC_KEY=<YOUR_LANGFUSE_PUBLIC_KEY>",
"-e",
"LANGFUSE_SECRET_KEY=<YOUR_LANGFUSE_SECRET_KEY>",
"-e",
"LANGFUSE_BASE_URL=<YOUR_LANGFUSE_BASE_URL>",
"ghcr.io/rodrigorjsf/langfuse-api-mcp:0.2.0"
],
"env": {
"LANGFUSE_PUBLIC_KEY": "<YOUR_LANGFUSE_PUBLIC_KEY>",
"LANGFUSE_SECRET_KEY": "<YOUR_LANGFUSE_SECRET_KEY>",
"LANGFUSE_BASE_URL": "<YOUR_LANGFUSE_BASE_URL>"
}
}
}
}Let your agent find its tools
Find servers like Langfuse API MCP from your agent. Add Koskamo once to search the index for endpoints, access requirements, and client setup. Public discovery needs no key.