# AgentOps-AI/agentops

Python SDK for AI agent monitoring, LLM cost tracking, benchmarking, and more. Integrates with most LLMs and agent frameworks including CrewAI, Agno, OpenAI Agents SDK, Langchain, Autogen, AG2, and CamelAI

Repository: https://github.com/AgentOps-AI/agentops
Canonical: https://ross.abutalabs.com/products/agentops
Homepage: https://agentops.ai
Language: Python
License: MIT
License Family: permissive
Topics: agent, agentops, ai, evals, evaluation-metrics, llm, anthropic, autogen, cost-estimation, crewai, groq, langchain, mistral, ollama, openai, agents-sdk, openai-agents
Last push: 2026-06-25T08:25:03+00:00

## Health v2 (maintenance only)
Score: 72/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 89, release rhythm 45, longevity 79
- inputs: {"age_days": 1114, "days_push": 69, "days_rel": 369, "gap_med": 3.0, "n_releases_24m": 49}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 5797, forks 615 (observed 2026-08-28T04:09:29.703256+00:00)

## What it is
AgentOps is a Python SDK and observability platform for monitoring, debugging, and evaluating AI agents and LLM applications. It auto-instruments LLM calls, tool usage, and multi-agent interactions, tracking traces, costs, and errors with integrations for frameworks like CrewAI, LangChain, AutoGen, and the OpenAI Agents SDK.

## Use cases
- track llm costs per agent run
- debug ai agent workflows with session replay
- monitor ai agents in production
- evaluate and benchmark llm agents
- trace multi-agent interactions in crewai or langchain
- audit agent logs and errors

## When to choose
- you build AI agents with frameworks like CrewAI, LangChain, AutoGen, or OpenAI Agents SDK and need observability
- you want LLM cost tracking and evaluation metrics with minimal code (two lines to init)
- you need session replay and time-travel debugging for agent runs

## When to avoid
- you need general-purpose APM for non-LLM applications
- you want fully self-hosted observability without the AgentOps dashboard service
- you use a stack outside Python/TypeScript SDKs or unsupported LLM providers

## Facets
- artifact type: library
- maturity: active
- function: monitoring, tracing, benchmarking, analytics, logging, sdk
- domain: large-language-models, developer-tools, monitoring, machine-learning
- platform: python, cross-platform
- tags: llm-observability, agent-monitoring, cost-tracking, llm-tracing, agent-evaluation, session-replay, ai-agents, nodejs

## Member repositories
- AgentOps-AI/agentops (main) score 72

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:29.703256+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:52:46.909551+00:00, confidence not recorded.
  - readme: https://github.com/AgentOps-AI/agentops (fetched 2026-08-28T04:09:29.703256+00:00, sha f797bb1a685d)
  - homepage: https://agentops.ai (fetched 2026-08-29T08:47:59.979468+00:00, sha cf2bd381c35b)
  - site_page: https://docs.agentops.ai (fetched 2026-08-29T08:47:59.988322+00:00, sha 1caf236089e3)
  - registry_pypi: https://pypi.org/pypi/agentops/json (fetched 2026-08-29T08:47:59.990289+00:00, sha 1738bb4701d6)
- Data as of 2026-08-30T08:39:29.467469+00:00.