"Understanding RAG (Retrieval-Augmented Generation or related topics)" matching MCP servers:

A black-box flight recorder for RAG retrieval inside MCP agents. Logs what chunks the model saw, scores, sources, and rankings - so you can audit, replay, and diff retrieval runs after the fact.

A modular Retrieval-Augmented Generation (RAG) framework that provides hybrid search and knowledge retrieval capabilities via the Model Context Protocol. It enables users to integrate document-based knowledge into LLM workflows with support for dense/sparse retrieval, reranking, and observability.

Exposes knowledge retrieval to AI tools over MCP with dense, sparse, and hybrid search, RRF fusion, configurable reranking, and modular ingestion of PDF/Markdown documents including image captioning. It also provides query tracing and Ragas-based evaluation so pipelines can be swapped, debugged, and tested independently.

这是面向智能运维场景的 MCP 工具服务，让 Agent 能够按需检索日志内容并获取主机与应用的监控指标数据。它把实时的日志与指标证据接入 Plan-Execute-Replan 诊断流程，从而支撑自动故障排查与根因分析。

Enables cloud and platform engineers to investigate AWS incidents by grounding responses in approved runbooks and operational evidence, with authenticated access, citations, observability, and human approval before any write action.

Enables read-only inspection of local coding-agent sessions across Claude Code, OpenCode, Cursor, and Codex, including running status, topics, todos, and transcripts, without writing or sending messages.

Enables MCP-capable clients to monitor and manage Apache Kafka clusters, including topics, partitions, configs, and consumer groups with lag, while enforcing configurable security modes, topic allowlists, protected topics, and audit logging.

Enables AI assistants to interact with tshark and related command-line tools for live packet capture, PCAP analysis, filtering, stream following, and capture file merging.

Enables plant operators to query alarm data, asset information, trends, correlations, and operator recommendations through 14 tools, forming a core part of a multi-MCP enterprise operations copilot.

Enables MCP-compatible clients to access model registry, prediction logging, drift detection, bias/fairness auditing, compliance and governance queries, explanation generation, LLM/RAG monitoring, content safety, alert management, multi-agent workflow tracking, and performance metrics through 197 tools across 18 domains. Supports stdio and hosted Streamable HTTP transports, plus MCP Apps interactive audit cards.

Enables LLM evaluation and observability by uploading documents, building test sets, running RAG pipelines, and automatically scoring answers for groundedness, hallucination risk, retrieval quality, latency, and cost, with tools exposed to MCP-compatible clients.

An extensible MCP hub that exposes internal services (chat, observability, RAG) as namespaced tools via FastMCP, with OpenAPI auto-generation, auth, and resilient error handling.

Provides cryptographically signed, tamper-evident compute usage receipts via a Model Context Protocol server, exposing tools to record usage, verify receipts, list the ledger, and verify ledger integrity without requiring any account, API key, or single cloud provider.

Enables classification of GPU workloads as training, inference, or idle from telemetry data, with tools for one-shot classification, benchmarking, and audit log verification.

16-tool MCP server for Google PageSpeed Insights & Chrome UX Report APIs. Analyze, compare, and optimize web performance directly through Claude, Cursor, or any MCP-compatible AI client.

Read-only MCP server for Grafana Loki: lets AI agents (Claude Code, Cursor, Copilot) discover labels, read bounded LogQL pages, count lines by label or time, and export logs — across several Loki stands

MCP server exposing rlg (RustLogs) log files as tools for on-call / SRE agent workflows. Three tools over JSON-RPC stdio — tail_log (last N events, filtered by level/target), filter_log (predicate filtering), summarize_errors (grouped error taxonomy). Pure Rust. MIT OR Apache-2.0.

Compares AI coding tool spend before and after a migration -- Cursor, Claude Code, GitHub Copilot, OpenCode, Codex CLI -- via an MCP server, real numbers from each vendor's own API or local logs.

Uptime monitoring for websites, APIs, SSL certificates, domain expiry, ping and TCP/UDP ports. 15 tools to list, create, pause and delete monitors, pull incident timelines with error codes, and read hourly or daily uptime and response-time statistics.

MCP server for the Chrome UX Report, providing real-user Core Web Vitals (LCP, INP, CLS) and historical trends for any origin or URL via natural language queries.

Enables AI assistants to search, summarize, and inspect Kibana/OpenSearch Dashboards logs through the user's logged-in browser session, eliminating the need for API keys or service accounts.

Enables AI agents to monitor server health and capacity, diagnose outages, inspect Docker and deployment status, and perform safe, bounded recovery actions via MCP without granting unrestricted shell or SSH access.

An MCP server that exposes a dbt project's run state as tools, enabling AI assistants to answer questions like 'is the warehouse healthy?' or 'what broke and why?' via plain English, using read-only access to dbt artifacts and warehouse data.

Enables coding agents to read performance findings from sixty, get detailed evidence for each finding, and close or resolve them with descriptions, while also checking service telemetry and providing installation instructions.

Laravel Telescope-style debugging for Hono apps whose dashboard endpoint is also an MCP server: read recent exceptions with the request that produced them, recent requests filtered by status or duration, one request in full, and the slowest queries. Five read-only tools, zero runtime dependencies.

MCP server for analyzing Perfetto traces with LLMs — query .pftrace files in PerfettoSQL via Claude Code or any MCP client

Enables AI agents to check the health of internet infrastructure and specific services, helping diagnose whether issues are local or external.

Enables AI clients to assess infrastructure-as-code deployment risk, generate incident response runbooks, review CI/CD delivery readiness, and calculate SLO error budgets through structured advisory tools. Runs locally over stdio without cloud credentials or external API calls, returning guidance for platform work, pull request review, and production readiness checks.

Zabbix MCP server and CLI in one Go binary. 15 task-shaped tools: current problems, host investigation, why an alert did or did not arrive, metrics, maintenance. Zabbix 6.4+.

This MCP server enables AI assistants to read PBX state—extensions, DID numbers, SIP trunk providers, tenants, and live statistics—from ICTPBX via the ICTCore REST API, without placing calls or changing configuration.