Unified Tracing for a Mixed Agent Stack: Why OpenTelemetry Beats SDK Lock-In
Swimming_Champion_75 · reddit · 2026-10-23
The author's team runs a mixed agent stack—LangGraph, OpenAI Agents SDK, and some Vercel AI—and observability platforms only work well when you're all-in on their SDK, making useful traces cost a full rewrite.
The realization: OpenTelemetry is the way out. If a platform ingests OTEL traces, your framework matters much less—instrument each app and export spans in a common format.
The other half nobody discusses until it bites: prompt versioning. When an agent misbehaves in prod, the first question is which prompt version and model produced it; you need versioned deployments you can diff and roll back without redeploying everything.
Quick survey of options:
- LangSmith: smoothest if you're already a LangChain/LangGraph shop; supports OTEL.
- Langfuse: solid tracing, supports OTEL.
- Orq AI: documented OTEL integrations for LangGraph, Vercel AI, Pydantic AI, OpenAI Agents SDK, and Agno; version history with diff/restore, head-to-head prompt experiments. Caveat: hosted SaaS runs on EU infrastructure.
The author asks the community whether to go all-in on OTEL, standardize one SDK, or keep stitching per agent.
More from coding & agent
- OpenJet claims Qwen3.8-27B coding runs 1.9x faster than llama.cpp on one RTX 3090 — JeremyCMorgan · 2026-10-23
- Building an autonomous SEO agent in Hermes with three skills: programmatic pages, Keywords Everywhere, DataForSEO — EXM7777 · 2026-10-23
- Building a sports betting AI agent with onchain execution and human approval by default: architecture and lessons — Dry-Comfortable-2678 · 2026-10-23
- Capy 0.5.0 lets Claude Pro/Max subscribers reuse their plan quota, adds OIDC cloud access — garrytan · 2026-10-23
- You.com CEO Richard Socher: legacy code is human slop, let AI clean it up — RichardSocher · 2026-10-23
- Let agents design a single-line glance component so you can track them from the sidebar — blixt · 2026-10-23