- CollectionObservability & Tracing
- Classificationagent-native
- MCPUnknown
- Agent SkillUnavailable
- InterfacesMCP · REST · SDK · http · API
What it does
AgentOps instruments AI agents and LLM apps with sessions, waterfalls of LLM/tool events, token/cost breakdowns, and framework integrations (LangChain, CrewAI, AutoGen, OpenAI Agents SDK, Google ADK, Smolagents, Haystack, AG2, etc.). It is explicitly positioned for testing and debugging agents, with a public API to pull trace data into other systems.
Agent-facing interfaces
- Interaction pattern: SDK auto-instrumentation + Public REST API + MCP docs helper
- Catalog protocols: MCP, REST, SDK, http, API
- MCP: ⚠️ Docs MCP for Mintlify (chat with documentation) — not a trace-ingestion MCP. Trace data flows through SDK and REST.
- Agent Skill: ⚠️ No official npx skills add agentops/… pack documented as primary.
- Onboarding type:
mcp
How agents onboard
Interaction pattern: SDK auto-instrumentation + Public REST API + MCP docs helper
import agentops
agentops.init("<YOUR API KEY HERE>")
See the full onboarding section in the dossier.
Protocol surface
| Interface | Detail |
|---|---|
| Python SDK | pip install agentops |
| TypeScript SDK | See docs |
| REST | https://api.agentops.ai |