Observability for developers building connectors

2026-06-08 · Source: Claude Blog · Field: Technology & Digital — Software Development & Engineering, Artificial Intelligence & Machine Learning, Data Science & Analytics · Depth: Intermediate, quick

Summary

Claude has introduced new observability features and in-app submission capabilities for developers building connectors for its product ecosystem. Now available in public beta, a dedicated dashboard allows owners of published connectors to monitor performance across Claude, Claude Code, and Cowork. This includes tracking adoption metrics like active users, total tool calls, and directory rank over time. Developers can also diagnose issues by viewing health scores, error rates, and latency at a glance, with detailed per-tool error breakdowns. Furthermore, the dashboard provides usage breakdowns by product to understand user engagement patterns. Access to these features requires Admin or Owner permissions on Team or Enterprise plans, with Enterprise Owners able to delegate access through custom roles. Developers can also now submit their Model Context Protocol (MCP) servers directly within Claude to join the directory, which currently hosts over 300 third-party connectors.

Key takeaway

For connector developers managing and optimizing integrations for Claude, these new observability tools are crucial. You can now directly monitor your connector's adoption, diagnose performance issues with detailed error rates and latency metrics, and understand user engagement across Claude products. This integrated visibility allows you to proactively improve connector health and drive adoption, ensuring your Model Context Protocol (MCP) servers perform optimally within the Claude ecosystem. Utilize the in-app submission to streamline directory inclusion.

Key insights

Claude now offers integrated observability and direct submission for connector developers to enhance performance and adoption.

Principles

• Connector performance is measurable via adoption, health, and usage.
• Granular error breakdowns aid efficient debugging.
• Product-specific usage data informs engagement strategy.

Method

The article describes a dashboard for monitoring, debugging, and improving connectors. It involves tracking adoption, diagnosing errors and latency with per-tool breakdowns, and analyzing usage across Claude products.

In practice

• Monitor active users and tool calls.
• Pinpoint errors using per-tool breakdowns.
• Compare usage across Claude products.

Topics

• Claude Connectors
• Observability
• Performance Monitoring
• Model Context Protocol
• Developer Tools
• API Management

Best for: AI Architect, AI Product Manager, AI Engineer, MLOps Engineer

Related on AIssential

• The AWS API MCP Server connector for Claude: What It Is and How to Set It Up. — Natural language processing can significantly reduce friction in cloud resource management.
• Monitoring reliably at scale — Reliable observability requires isolating its components from the systems it monitors to prevent circular dependencies and ensure visibility during outages.
• v2.1.86 — Recent Claude Code updates enhance performance, fix bugs, and introduce new features for improved developer workflows.
• Mind the Gap (In your Agent Observability) — Amy Boyd & Nitya Narasimhan, Microsoft — Effective AI agent observability requires continuous evaluation, monitoring, and automated optimization to manage non-determinism and ensure reliability.
• It just got easier for Claude to check in on your WordPress site — WordPress now integrates with Claude, offering read-only data access for site analysis and management.
• Building Observability for a Production GenAI System: An Internal Knowledge Base End-to-End — Integrated observability across cost, quality, and latency is critical for production GenAI systems.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Claude Blog.