The open-source, local-first alternative to LangSmith. No cloud. No setup.
I built AgentWatch to solve a problem I had: AI agents burning tokens and money behind my back. Now I can monitor everything in real-time. Feedback welcome! 🤖💸
Drop-in observability for LLM apps — quality scoring, hallucination detection, cost tracking and agent debugging
Decision-quality observability for LLM agents — scores whether each decision was good, not just how fast it ran. Tool ROI, reasoning coherence, confidence signals. Pure Node.js, zero frameworks.
This repo contains detailed notes on LangSmith concepts including traces, runs, observability, and integrations with LangChain, RAG, and LangGraph.
🟢 No LangSmith required. DeerFlow Trace Edition — super agent harness (fork of bytedance/deer-flow) with BUILT-IN Execution Flow trace panel. Every LangGraph checkpoint, sub-agent, tool, and model node is inspectable directly inside chat UI. 100% local, zero SaaS, zero per-trace fees, zero data leaving your machine.
Hand-built LLM agent observability: trace logging, token budget/sliding window, and LLM-as-judge scoring — no external SaaS. Wraps any agent function with timing, context truncation, and semantic (not string-match) answer evaluation.
Self-hosted observability and debugging platform for AI agents with traces, execution DAGs, cost analytics, latency stats, and live events.
Official TypeScript/Node SDK for Observyze. 1-line auto-instrumentation for OpenAI, Anthropic, Gemini, Vercel AI & LangChain, with OpenTelemetry tracing, execution budgets, PII redaction & guardrails.
To associate your repository with the langsmith-alternative topic, visit your repo's landing page and select "manage topics."