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.
Lets an AI client query agent runs, failure signals, health scores, and fixes from Dunetrace inside the editor.
Monitors MCP traffic in real time and detects command injection, PII leaks, data exfiltration, and tool poisoning threats.
Analyzes PostgreSQL bloat, identifies vacuum blockers, and detects bloated or unused indexes with actionable repair instructions.
$claude mcp add pg-bloat-detective -- uvx pg-bloat-detective
Connects an openITCOCKPIT monitoring instance to an LLM to analyze host states, service health, downtimes, and automate monitoring configurations.
$claude mcp add openitcockpit-mcp-server -- docker run -i --rm docker.io/openitcockpit/mcp-server:0.5.0
Lets coding agents inspect recorded browser flows with network, console, storage, DOM, and screen data for debugging and regression testing.
Lets an AI client capture live traffic and analyze pcap files with tshark, applying filters, following streams, and exporting JSON.
Monitors keyword momentum across platforms like Google, TikTok, and Reddit to send automated breakout alerts via Slack, Discord, or email.
Lets AI clients triage, investigate, and operate Icinga installations through Icinga 2 Core, Icinga Web, and Icinga Director REST APIs.
Analyzes LLM spending and latency to provide actionable insights and model routing suggestions directly within AI clients.
Provides role-aware, evidence-backed log analysis tools for BGL, OpenStack, and Zookeeper datasets stored in SereneDB.
Records and debugs MCP agents by capturing LangGraph steps, model calls, and tool calls in a web-based UI.
Measures and tracks how often AI models recommend your product or library compared to competitors using automated scans and CI testing.
Records independent, dated records of AI agent and MCP server publications and provides signed receipts for verified facts.
$claude mcp add --transport http gaip-outcome-value https://www.gaipagents.com/mcp/agents/outcome-valueMaintains an independent, dated record of AI agents, MCP servers, and online shops to provide verifiable facts and signed receipts.
$claude mcp add --transport http gaip-listing-observer https://www.gaipagents.com/mcp/agents/listing-observerExposes NVIDIA GPU metrics like VRAM, utilization, and temperature as tools for AI clients to query real-time hardware status.
$claude mcp add mcp-gpu-server -- uvx mcp-gpu-server
Exposes NVIDIA GPU metrics like VRAM, utilization, and temperature as tools for AI clients.
$claude mcp add mcp-gpu-server -- uvx gpu-server
Tracks and attributes AI agent costs across multiple vendors by monitoring token usage directly within your process.
$claude mcp add focxle-python -- uvx focxle
Analyze MCP server usage, investigate sessions, and instrument MCP servers for observability.
$claude mcp add --transport http plugins https://api.flowlines.ai/mcp
Provides turn-level governance observability and token flow analysis for agent harness transcripts.
$claude mcp add harness-scope -- npx -y hscope
Tracks news article mention volume trends and growth for specific topics without web scraping.
$claude mcp add --transport http news-trends-api https://news-volume.api.trendsapi.ai/mcp