Monitoring and observability MCP servers for logs, metrics, traces and error tracking: let an AI assistant investigate incidents with real data. 699 servers in this category.
Enables AI assistants to monitor, troubleshoot, and predict issues across Kubernetes, OpenShift, and Tekton environments using specialized SRE tools.
$claude mcp add lumino-mcp-server -- uvx lumino-mcp-server
Lets an AI client analyze vibration signals to detect machine faults, assess ISO severity, and generate diagnostic reports.
Lets AI agents query Langfuse traces, sessions, exceptions, prompts, datasets, and metrics for debugging and observability.
$claude mcp add langfuse-mcp -- uvx langfuse-mcp
Lets an AI client query alerts, silences, receivers, and alert groups in Prometheus Alertmanager and create or delete silences and alerts.
Lets an AI client run read-only SQL queries, inspect schemas, and analyze MySQL server performance, health, and security.
Lets an AI client debug Zuul CI build failures, search logs, inspect pipelines and jobs, and monitor infrastructure.
$claude mcp add mcp-zuul -- uvx mcp-zuul
Manage Proxmox VE clusters from an AI client: list nodes, provision VMs and containers, run backups, snapshots, and migrations.
$claude mcp add mcp-proxmox -- uvx mcp-proxmox
Exposes Microsoft Clarity's Data Export API to allow AI agents to analyze behavioral UX metrics like rage clicks, dead clicks, and JavaScript errors.
$claude mcp add microsoft-clarity-mcp -- uvx microsoft-clarity-mcp
Allows an AI assistant to monitor WiFi temperature sensors, check freezer status, and retrieve humidity, battery, and alert data.
$claude mcp add tempstick-mcp -- uvx tempstick-mcp
Allows AI agents to query Prometheus metrics using PromQL, inspect active alerts, and explore scrape targets in a read-only environment.
$claude mcp add prometheus-mcp -- uvx prometheus-mcp
Monitors web page changes using noise-filtered deterministic diffs and SHA-256 cryptographic evidence.
$claude mcp add page-monitor-mcp-server -- docker run -i --rm ghcr.io/jhfg31410-ai/page-monitor-mcp-server:0.1.0
Lets an AI client diagnose and monitor a self-hosted media stack including Sonarr, Radarr, Jellyfin, and Docker containers.
$claude mcp add oraclarr -- uvx oraclarr-mcp
Monitors the health of local AI stacks by checking Ollama nodes, NVIDIA GPUs, and tailnet peers with graded confidence findings.
$claude mcp add stack-guardian -- uvx stack-guardian
Enables AI clients to perform read-only packet capture analysis using tshark across telecom and network protocols like Diameter, GTP, SIP, and RTP.
$claude mcp add tshark-mcp -- uvx tshark-mcp-server
Analyzes Claude Code transcripts to measure token efficiency and the cost-weighted share of useful model output versus context re-reading.
$claude mcp add contextburn -- uvx contextburn
Analyzes local Codex session metadata to provide exact facts about token usage, costs, credits, and tool patterns.
Allows AI agents to monitor website changes, manage watch lists, and retrieve historical snapshots and diffs from a ChangeDetection.io instance.
$claude mcp add changedetection-mcp -- uvx changedetection-mcp
Allows an AI agent to investigate Google Cloud incidents, take resource inventory, and identify cost-optimization opportunities using read-only access.
$claude mcp add platform-mcp -- uvx platform-mcp
Extracts representative log snippets and Python tracebacks using Drain3 clustering to reduce large logs for AI analysis.
$claude mcp add logdetective-mcp -- uvx logdetective-mcp
Analyzes web page performance using Google PageSpeed Insights and Chrome UX Report to provide median scores and uncertainty spreads across multiple runs.
$https://github.com/Considus/pagespeed-insights-mcp/releases/download/v2.0.1/pagespeed-insights-mcp-2.0.1.mcpb