LA

langfuse-migration-deep-dive

A guide for migrating data and platforms between Langfuse instances or other observability tools.

Install

mkdir -p .claude/skills/langfuse-migration-deep-dive && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/9310" && unzip -o skill.zip -d .claude/skills/langfuse-migration-deep-dive && rm skill.zip

Installs to .claude/skills/langfuse-migration-deep-dive

Activation

This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.

Execute complex Langfuse migrations including data migration and platform
73 charsno explicit “when” trigger
Advanced

Key capabilities

  • Export traces, scores, prompts, and datasets from source instances
  • Import data into target Langfuse instances
  • Validate migration success by comparing trace counts
  • Implement dual-write patterns for zero-downtime migration
  • Execute rollback procedures to source instances

How it works

The skill utilizes the Langfuse client API to programmatically export data from a source instance and import it into a target instance. It includes validation scripts to ensure data parity and dual-write patterns to maintain observability during the transition.

Inputs & outputs

You give it
Source and target Langfuse API credentials
You get back
Migrated data and validation report

When to use langfuse-migration-deep-dive

  • Migrate from LangSmith to Langfuse
  • Move Langfuse from cloud to self-hosted
  • Perform cross-instance data migration
  • Implement zero-downtime dual-write patterns

About this skill

Langfuse Migration Deep Dive

Current State

!npm list langfuse @langfuse/client 2>/dev/null | head -5 || echo 'No langfuse packages'

Overview

Comprehensive guide for complex migrations: cloud-to-self-hosted, LangSmith-to-Langfuse, cross-instance data migration, and zero-downtime dual-write patterns.

Prerequisites

  • Understanding of source and target Langfuse instances
  • API keys for both source and target
  • Git branch for migration work
  • Rollback plan documented

Migration Scenarios

ScenarioComplexityDowntimeData Loss Risk
Cloud to Cloud (different project)LowNoneNone
Cloud to Self-hostedMediumMinutesLow
Self-hosted to CloudMediumMinutesLow
LangSmith to LangfuseHighHoursMedium
SDK v3 to v4+ (no data migration)LowNoneNone

Instructions

Step 1: Export Data from Source Instance

// scripts/export-langfuse.ts
import { LangfuseClient } from "@langfuse/client";
import { writeFileSync, mkdirSync } from "fs";

const source = new LangfuseClient({
  publicKey: process.env.SOURCE_LANGFUSE_PUBLIC_KEY,
  secretKey: process.env.SOURCE_LANGFUSE_SECRET_KEY,
  baseUrl: process.env.SOURCE_LANGFUSE_BASE_URL,
});

async function exportAll(outputDir: string) {
  mkdirSync(outputDir, { recursive: true });

  // Export traces
  let page = 1;
  let allTraces: any[] = [];
  let hasMore = true;

  console.log("Exporting traces...");
  while (hasMore) {
    const result = await source.api.traces.list({ limit: 100, page });
    allTraces.push(...result.data);
    hasMore = result.data.length === 100;
    page++;
    await new Promise((r) => setTimeout(r, 200)); // Rate limit
  }
  writeFileSync(`${outputDir}/traces.json`, JSON.stringify(allTraces, null, 2));
  console.log(`  Exported ${allTraces.length} traces`);

  // Export scores
  page = 1;
  let allScores: any[] = [];
  hasMore = true;

  console.log("Exporting scores...");
  while (hasMore) {
    const result = await source.api.scores.list({ limit: 100, page });
    allScores.push(...result.data);
    hasMore = result.data.length === 100;
    page++;
    await new Promise((r) => setTimeout(r, 200));
  }
  writeFileSync(`${outputDir}/scores.json`, JSON.stringify(allScores, null, 2));
  console.log(`  Exported ${allScores.length} scores`);

  // Export prompts
  console.log("Exporting prompts...");
  const prompts = await source.api.prompts.list({ limit: 100 });
  writeFileSync(`${outputDir}/prompts.json`, JSON.stringify(prompts.data, null, 2));
  console.log(`  Exported ${prompts.data.length} prompts`);

  // Export datasets
  console.log("Exporting datasets...");
  const datasets = await source.api.datasets.list({ limit: 100 });
  const fullDatasets = [];
  for (const ds of datasets.data) {
    const items = await source.api.datasetItems.list({ datasetName: ds.name, limit: 1000 });
    fullDatasets.push({ ...ds, items: items.data });
    await new Promise((r) => setTimeout(r, 200));
  }
  writeFileSync(`${outputDir}/datasets.json`, JSON.stringify(fullDatasets, null, 2));
  console.log(`  Exported ${fullDatasets.length} datasets`);
}

exportAll("./migration-export");

Step 2: Import Data to Target Instance

// scripts/import-langfuse.ts
import { LangfuseClient } from "@langfuse/client";
import { readFileSync } from "fs";

const target = new LangfuseClient({
  publicKey: process.env.TARGET_LANGFUSE_PUBLIC_KEY,
  secretKey: process.env.TARGET_LANGFUSE_SECRET_KEY,
  baseUrl: process.env.TARGET_LANGFUSE_BASE_URL,
});

async function importAll(inputDir: string) {
  // Import prompts first (no dependencies)
  console.log("Importing prompts...");
  const prompts = JSON.parse(readFileSync(`${inputDir}/prompts.json`, "utf-8"));
  for (const prompt of prompts) {
    await target.api.prompts.create({
      name: prompt.name,
      prompt: prompt.prompt,
      type: prompt.type,
      config: prompt.config,
      labels: prompt.labels,
    });
    console.log(`  Imported prompt: ${prompt.name}`);
    await new Promise((r) => setTimeout(r, 100));
  }

  // Import datasets
  console.log("Importing datasets...");
  const datasets = JSON.parse(readFileSync(`${inputDir}/datasets.json`, "utf-8"));
  for (const ds of datasets) {
    await target.api.datasets.create({
      name: ds.name,
      description: ds.description,
      metadata: { ...ds.metadata, migratedFrom: "source-instance" },
    });

    for (const item of ds.items || []) {
      await target.api.datasetItems.create({
        datasetName: ds.name,
        input: item.input,
        expectedOutput: item.expectedOutput,
        metadata: item.metadata,
      });
      await new Promise((r) => setTimeout(r, 50));
    }
    console.log(`  Imported dataset: ${ds.name} (${ds.items?.length || 0} items)`);
  }

  console.log("Import complete.");
  console.log("Note: Traces and scores are historical -- they reference old trace IDs.");
  console.log("New traces will be created by your application pointing to the target.");
}

importAll("./migration-export");

Step 3: Dual-Write for Zero-Downtime Migration

Write traces to both instances during transition:

// src/lib/dual-write-langfuse.ts
import { LangfuseSpanProcessor } from "@langfuse/otel";
import { NodeSDK } from "@opentelemetry/sdk-node";

// Create processors for both instances
const sourceProcessor = new LangfuseSpanProcessor({
  publicKey: process.env.SOURCE_LANGFUSE_PUBLIC_KEY,
  secretKey: process.env.SOURCE_LANGFUSE_SECRET_KEY,
  baseUrl: process.env.SOURCE_LANGFUSE_BASE_URL,
});

const targetProcessor = new LangfuseSpanProcessor({
  publicKey: process.env.TARGET_LANGFUSE_PUBLIC_KEY,
  secretKey: process.env.TARGET_LANGFUSE_SECRET_KEY,
  baseUrl: process.env.TARGET_LANGFUSE_BASE_URL,
});

// Both processors receive all spans
const sdk = new NodeSDK({
  spanProcessors: [sourceProcessor, targetProcessor],
});
sdk.start();

// Migration timeline:
// Week 1: Dual-write enabled, verify target receives traces
// Week 2: Validate data parity between instances
// Week 3: Switch primary to target, keep source as backup
// Week 4: Remove source processor

Step 4: Validate Migration

// scripts/validate-migration.ts
import { LangfuseClient } from "@langfuse/client";

const source = new LangfuseClient({
  publicKey: process.env.SOURCE_LANGFUSE_PUBLIC_KEY,
  secretKey: process.env.SOURCE_LANGFUSE_SECRET_KEY,
  baseUrl: process.env.SOURCE_LANGFUSE_BASE_URL,
});

const target = new LangfuseClient({
  publicKey: process.env.TARGET_LANGFUSE_PUBLIC_KEY,
  secretKey: process.env.TARGET_LANGFUSE_SECRET_KEY,
  baseUrl: process.env.TARGET_LANGFUSE_BASE_URL,
});

async function validate() {
  // Compare prompt counts
  const sourcePrompts = await source.api.prompts.list({ limit: 100 });
  const targetPrompts = await target.api.prompts.list({ limit: 100 });
  console.log(`Prompts: source=${sourcePrompts.data.length}, target=${targetPrompts.data.length}`);

  // Compare dataset counts
  const sourceDatasets = await source.api.datasets.list({ limit: 100 });
  const targetDatasets = await target.api.datasets.list({ limit: 100 });
  console.log(`Datasets: source=${sourceDatasets.data.length}, target=${targetDatasets.data.length}`);

  // Compare recent trace counts (dual-write period)
  const since = new Date(Date.now() - 3600000).toISOString();
  const sourceTraces = await source.api.traces.list({ fromTimestamp: since, limit: 100 });
  const targetTraces = await target.api.traces.list({ fromTimestamp: since, limit: 100 });
  console.log(`Recent traces (1h): source=${sourceTraces.data.length}, target=${targetTraces.data.length}`);

  const variance = Math.abs(sourceTraces.data.length - targetTraces.data.length) / Math.max(sourceTraces.data.length, 1);
  console.log(`Trace variance: ${(variance * 100).toFixed(1)}% (target: <5%)`);
}

validate();

Step 5: Cutover and Cleanup

// After validation passes:

// 1. Update environment variables to point to target
// LANGFUSE_PUBLIC_KEY=pk-lf-target-...
// LANGFUSE_SECRET_KEY=sk-lf-target-...
// LANGFUSE_BASE_URL=https://target.langfuse.com

// 2. Remove dual-write (use single processor)
const sdk = new NodeSDK({
  spanProcessors: [targetProcessor], // Only target
});

// 3. Keep source instance running for 30 days (rollback window)
// 4. After 30 days, decommission source

Rollback Plan

set -euo pipefail
# If migration fails, switch back to source:

# 1. Update environment variables
export LANGFUSE_PUBLIC_KEY="pk-lf-source-..."
export LANGFUSE_SECRET_KEY="sk-lf-source-..."
export LANGFUSE_BASE_URL="https://source.langfuse.com"

# 2. Restart application
# 3. Verify traces flowing to source

Error Handling

IssueCauseSolution
Export timeoutToo much dataPaginate with smaller page sizes
Import duplicatesRe-running importUse idempotent creates with unique names
Dual-write divergenceOne instance failingMonitor both, alert on variance > 5%
Missing promptsNot exportedExport prompts before datasets

Resources

When not to use it

  • When migrating non-Langfuse observability data
  • When the source and target instances have incompatible schemas

Prerequisites

API keys for source and target instancesGit branch for migration workDocumented rollback plan

Limitations

  • Traces and scores are historical and reference old trace IDs
  • Export may timeout with large datasets

How it compares

This approach uses automated scripts for data transfer, whereas manual migration is prone to data loss and configuration errors.

Compared to similar skills

langfuse-migration-deep-dive side by side with the closest alternatives in the catalog.

SkillInstallsUpdatedSafetyDifficulty
langfuse-migration-deep-dive (this skill)024dReviewAdvanced
distributed-tracing52moNo flagsIntermediate
service-mesh-observability52moNo flagsAdvanced
terminal-context57moReviewAdvanced

Try saying

Example prompts that trigger this skill in your AI assistant.

More by jeremylongshore

View all by jeremylongshore

analyzing-logs

jeremylongshore

Analyze application logs to detect performance issues, identify error patterns, and improve stability by extracting key insights.

14123

ollama-setup

jeremylongshore

Configure auto-configure Ollama when user needs local LLM deployment, free AI alternatives, or wants to eliminate hosted API costs. Trigger phrases: "install ollama", "local AI", "free LLM", "self-hosted AI", "replace OpenAI", "no API costs". Use when appropriate context detected. Trigger with relevant phrases based on skill purpose.

1167

backtesting-trading-strategies

jeremylongshore

Backtest crypto and traditional trading strategies against historical data. Calculates performance metrics (Sharpe, Sortino, max drawdown), generates equity curves, and optimizes strategy parameters. Use when user wants to test a trading strategy, validate signals, or compare approaches. Trigger with phrases like "backtest strategy", "test trading strategy", "historical performance", "simulate trades", "optimize parameters", or "validate signals".

1071

generating-database-seed-data

jeremylongshore

Process this skill enables AI assistant to generate realistic test data and database seed scripts for development and testing environments. it uses faker libraries to create realistic data, maintains relational integrity, and allows configurable data volumes. u... Use when working with databases or data models. Trigger with phrases like 'database', 'query', or 'schema'.

1033

cursor-codebase-indexing

jeremylongshore

Execute set up and optimize Cursor codebase indexing. Triggers on "cursor index setup", "codebase indexing", "index codebase", "cursor semantic search". Use when working with cursor codebase indexing functionality. Trigger with phrases like "cursor codebase indexing", "cursor indexing", "cursor".

885

testing-mobile-apps

jeremylongshore

Execute mobile app testing on iOS and Android devices/simulators. Use when performing specialized testing. Trigger with phrases like "test mobile app", "run iOS tests", or "validate Android functionality".

810

You might also like

distributed-tracing

wshobson

Implement distributed tracing with Jaeger and Tempo to track requests across microservices and identify performance bottlenecks. Use when debugging microservices, analyzing request flows, or implementing observability for distributed systems.

577

service-mesh-observability

wshobson

Implement comprehensive observability for service meshes including distributed tracing, metrics, and visualization. Use when setting up mesh monitoring, debugging latency issues, or implementing SLOs for service communication.

574

terminal-context

aelaguiz

Complete Kitty terminal awareness + control for coding agents: list panes/tabs, read scrollback, map ports→processes, parse per-pane git/last-command metadata (from shell hooks), and send commands/focus panes. Use when user mentions "another terminal", "is the server running", "what failed", or you need to run/inspect commands across panes.

567

observability-engineer

sickn33

Build production-ready monitoring, logging, and tracing systems. Implements comprehensive observability strategies, SLI/SLO management, and incident response workflows. Use PROACTIVELY for monitoring infrastructure, performance optimization, or production reliability.

1242

prometheus-configuration

wshobson

Set up Prometheus for comprehensive metric collection, storage, and monitoring of infrastructure and applications. Use when implementing metrics collection, setting up monitoring infrastructure, or configuring alerting systems.

645

langfuse

davila7

Expert in Langfuse - the open-source LLM observability platform. Covers tracing, prompt management, evaluation, datasets, and integration with LangChain, LlamaIndex, and OpenAI. Essential for debugging, monitoring, and improving LLM applications in production. Use when: langfuse, llm observability, llm tracing, prompt management, llm evaluation.

743

Search skills

Search the agent skills registry