AgentOps 0.4.17 (The MCP Update) Changelog
MCP Server, LangGraph + Agno support, Graph View, and more

MCP (Model Context Protocol) Launch
Chat with your traces using the new MCP integration. Compatible with Cursor, Windsurf, Claude, and more.
LangGraph Integration
Connect AgentOps workflows to LangGraph for richer automation & visualization. Seamlessly integrate your existing LangGraph workflows with AgentOps monitoring and debugging capabilities.
Dynamic Trace Metadata
Edit and monitor traces on the fly with our new update_trace_metadata
function. Update trace information in real-time for better debugging and monitoring.
π Performance & Reliability
- Stronger concurrency handling for smoother parallel operations
- General performance boosts with quicker page loads
- Auto-save protection for better data integrity
π¨ UX & Interface
- Instrumentation & Agno UX upgrades for cleaner interface
- Enhanced dark-mode controls and multi-language support
- Interface personalization improvements
π API & Integration
- Public API launched & refined for external integrations
- Span resource attributes capped for leaner traces
- Higher-accuracy metrics and faster processing
- Greater stability for reliable operations
π§ Fixes & Improvements
- Prompt-array rendering fix: Glitch-free display of simple prompt arrays
- Oversized-chat truncation: Lightning-fast loading of very long conversations
- Legacy MCP cleanup: Retired old components for a slimmer, more maintainable codebase
π Documentation & Community
- Revamped notebooks & docs with clearer examples
- Fixed links and new session-management guides
- Smithery MCP badge: Gamified recognition of user achievements
You might also like reading:

Build your first agent using OpenAI's Agents SDK and AgentOpsAI
Build your first agent with less than 50 lines of code!

Building Multi-Agent workflows with OpenAIβs new Agents SDK and AgentOpsAI
What is OpenAI's Agents SDK?

AgentOps, the Best Tool for AutoGen Agent Observability
Have eyes on your autogen agentic application