Langfuse Python Packages

langfuse

🪢 Langfuse Python SDK

operonx

Workflow engine where ops can yield — same async DAG handles batch jobs and event-driven streaming pipelines (pipecat-style callbot, STT → LLM → TTS). Python frontend, drop-in Rust runtime; 2–3× faster on chains, 10–30× on production workloads. LLMs, agents, RAG, native compute as first-class ops.

agentic-coding-harness

Run, watch, and meter Claude Code, Codex CLI, Gemini CLI, OpenCode, and Kiro from one CLI. Normalized events, cache-aware token usage, cost accounting verified against each agent's own records, budget caps, OpenTelemetry gen_ai spans, Langfuse traces, and a live web dashboard + MCP server.

langfuse-mcp

A Model Context Protocol (MCP) server for Langfuse, enabling AI agents to query Langfuse trace data for enhanced debugging and observability

argus-obs-semconv

Production-grade observability and agentic AIOps for AI systems — OpenTelemetry-native across independent polyglot apps; GenAI semantic conventions with token/cost attribution; sub-2s detection on a hot path that skips the datastore; LLM agents that find root cause and notify. ClickHouse + Langfuse. 100% self-hosted — no vendor lock-in.

argus-obs-sdk

Production-grade observability and agentic AIOps for AI systems — OpenTelemetry-native across independent polyglot apps; GenAI semantic conventions with token/cost attribution; sub-2s detection on a hot path that skips the datastore; LLM agents that find root cause and notify. ClickHouse + Langfuse. 100% self-hosted — no vendor lock-in.

argus-obs-schemas

Production-grade observability and agentic AIOps for AI systems — OpenTelemetry-native across independent polyglot apps; GenAI semantic conventions with token/cost attribution; sub-2s detection on a hot path that skips the datastore; LLM agents that find root cause and notify. ClickHouse + Langfuse. 100% self-hosted — no vendor lock-in.

axonium

A production-grade Python SDK for llama-server that streamlines authentication, token rotation, observability, and PII masking—helping AI architects ship secure, traceable LLM systems with enterprise-ready guardrails.

dunetrace

The runtime reliability layer for AI agents.

idun-agent-schema

🟪 Open-source runtime that ships any LangGraph or Google ADK agent as a production-ready FastAPI service. Bundled , AG-UI copilotkit API, chat UI, 15+ guardrails, MCP, OpenTelemetry, OIDC. One pip install. Self-hosted, no vendor lock-in.

idun-agent-engine

🟪 Open-source runtime that ships any LangGraph or Google ADK agent as a production-ready FastAPI service. Bundled , AG-UI copilotkit API, chat UI, 15+ guardrails, MCP, OpenTelemetry, OIDC. One pip install. Self-hosted, no vendor lock-in.

financebench-rag-agent

Multi-agent LangGraph RAG for financial Q&A — 72.7% on FinanceBench under κ=0.932 calibrated judge. RBAC at the vector layer, multi-party HITL on high-stakes answers, self-hosted LLM observability. pip install financebench-rag-agent

shekel

LLM budget control and cost governance for AI agents. Python library for token budgets, usage limits and guardrails for OpenAI, Anthropic, LangChain, LangGraph and agentic systems.

claude-code-langfuse-hook

Track Claude Code CLI usage in Langfuse per-user, per-project traces with token counts and cost estimates. Self-hosted or cloud.

agent-panorama

See what your AI agents do, whether it's worth it, and what it costs - a manager-readable report + local dashboard from Langfuse/LangSmith traces (or a one-line live callback). Open source, runs locally.

forgejudge

Open, always-on leaderboard + CI gate for autonomous coding agents — every patch sandboxed, every run traced, every regression fails the build. $0 stack.

behavioral-memory

Behavioral Memory for Tool Orchestration — Validated execution traces as memory for MCP-based agent systems. Based on IEEE paper.

iflow-mcp-avivsinai-langfuse-mcp

A Model Context Protocol (MCP) server for Langfuse, enabling AI agents to query Langfuse trace data for enhanced debugging and observability

docket-runtime

Observability-platform-agnostic triage runtime for LLM agent traces: classifies traces against a YAML failure-mode taxonomy, clusters failures, and drafts issues into trackers for human review.

atap

Agent Trace Analysis Platform (atap) — pluggable pipeline for trace representation, analysis, error classification, failure attribution and recovery for LLM agents

context-profiler

Framework-agnostic profiler for LLM agent context windows

hummcode

The Hummingbird Coding Agent — minimalist, model-agnostic, and composable. Built from first principles with tree-based memory, surgical file edits, and a rich terminal UI.

observix

Provider-agnostic observability for Python and AI apps. Instrument once with @observe; export natively to Phoenix, Langfuse, MLflow, Datadog and any OTLP backend simultaneously.

livingeval

Eval suites that tell you when they've gone blind: coverage, detection power and judge depth for LLM agent evaluation.