MCP servers · least privilege

Read-only MCP servers

1,620 MCP servers declare tools that only read: none of their tools is classed write or destructive. That makes them the safest place to start giving an agent access. Read-only describes the tools, not the code, so the audit verdict still decides whether to connect one.

Ask an AI about this pageGChatGPTCClaudePPerplexityGeGeminiXGrok
More
Cite

Cite this page

ModelsAtlas. “Read-only MCP servers”. Retrieved 29 Oct 2026. https://modelsatlas.com/mcp/read-only/page/44

Verdicts · 1,620 servers

✓SAFE913!CAUTION261✕BLOCK446

Transports

A server can list more than one transport.

1,620 servers

1,033–1,056 of 1,620 · ranked by trust score, then stars

Pg Aiguide

timescale/pg-aiguide

!CAUTION B

MCP server and Claude plugin for Postgres skills and documentation. Helps AI coding tools generate better PostgreSQL code.

★ 1,8491 readstreamable-http

LycheeMem

lycheemem/lycheemem

!CAUTION B

Lightweight Long-Term Memory for LLM Agents.

★ 1,0924 read

Opik

comet-ml/opik-mcp

!CAUTION B

Model Context Protocol (MCP) server for Opik, the open-source LLM observability and evaluation platform, built by Comet. Read traces, log scores, and manage prompts from Claude Code, Cursor, or VS Code.

★ 2192 readstdio

opik-mcp

comet-ml/opik-mcp

!CAUTION B

Model Context Protocol (MCP) server for Opik, the open-source LLM observability and evaluation platform, built by Comet. Read traces, log scores, and manage prompts from Claude Code, Cursor, or VS Code.

★ 2192 readstdio

ZeroSlop

manavmishra/ZeroSlop

!CAUTION B

Open-source Agent Skill that finds and removes AI slop from your writing

★ 1322 readstreamable-http

Next

vertile-ai/next-mcp-server

!CAUTION B

Help LLMs to understand your Next apps better

★ 994 readsse

Next.js Router Analyzer

jazelly/next-mcp-server

!CAUTION B

Help LLMs to understand your Next apps better

★ 994 readsse

Mindpilot

abrinsmead/mindpilot-mcp

!CAUTION B

See through your agent's eyes. Visualize legacy code, architect new systems, understand everything.

★ 902 readstdio

Multi-AI Advisor

yuchenssr/multi-ai-advisor-mcp

!CAUTION B

council of models for decision

★ 882 readstdio

SO-ARM100 Robot Control

ilialarchenko/robot_mcp

!CAUTION B

A simple MCP server for the SO-ARM100 control

★ 852 read

SAP Note Search

marianfoo/mcp-sap-notes

!CAUTION B

MCP Server to search and retrieve SAP Notes

★ 572 readstdio

TomTom

tomtom-international/tomtom-mcp

!CAUTION B

A Model Context Protocol (MCP) server providing TomTom's location services, search, routing, and traffic data to AI agents.

★ 5614 readstdio

Uptime Kuma

davidfuchs/mcp-uptime-kuma

!CAUTION B

A Model Context Protocol (MCP) server for Uptime Kuma version 2.

★ 5431 readstdio

One MCP

particlefuture/1mcpserver

!CAUTION B
★ 525 readstdio

TrueNAS Core

vespo92/truenascoremcp

✓SAFE B

A working TrueNAS Core MCP Server

★ 321 read

Chrome Google Search

cmann50/mcp-chrome-google-search

!CAUTION B
★ 252 readstdio

gitea-mcp-server

pxsloot/gitea-mcp-server

✓SAFE B

Gitea MCP Server – A Model Context Protocol implementation for Gitea. Provides tools and resources for AI agents to interact with Gitea repositories, issues, pull requests, and more, much more. Built with FastMCP for seamless Gitea integration.

★ 144 read

ProdKit

denisbilli/prodkit

✓SAFE B
★ 04 readstdio

agentic_services

impanyu/agentic_services

✓SAFE B

Paid claim verification with cited web evidence, source provenance, snapshots, and hashes.

★ 03 readstreamable-http

mailsift

itswl/mailsift

!CAUTION B

Read-only multi-mailbox triage: watch inboxes and spam, judge what actually matters with an LLM, and push only that to Feishu or a webhook. MCP server and a web view included.

★ 01 readstdio

popcorn-temporal-anchor

violetclaire/popcorn-temporal-anchor

!CAUTION B

Signed temporal evidence for agents: trusted timestamp and proof-of-existence verification, authorization expiry, and an RFC 3161 / OpenTimestamps comparison. Free offline verifier; evidence, not authorization or proof of execution.

★ 07 readstdio

BrowserTools MCP

AgentDeskAI/browser-tools-mcp

!CAUTION B

Monitor browser logs directly from Cursor and other MCP compatible IDEs.

★ 7,33015 readstdio

BrowserTools

agentdeskai/browser-tools-mcp

!CAUTION B

Monitor browser logs directly from Cursor and other MCP compatible IDEs.

★ 7,32615 readstdio

Interactive Feedback

minidoracat/mcp-feedback-enhanced

✓SAFE B

Enhanced MCP server for interactive user feedback and command execution in AI-assisted development, featuring dual interface support (Web UI and Desktop Application) with intelligent environment detection and cross-platform compatibility.

★ 3,7651 read

Other risk hubs

By job

Questions

What does read-only mean here?
A server is listed here when its source declares at least one tool and every tool is classed read: no write tools and no destructive tools.
Is a read-only server always safe?
No. 913 are SAFE, 261 CAUTION and 446 BLOCK on this list. A tool list says what the tools do, not whether the surrounding code is clean, so a read-only server can still be BLOCK.
Why are some servers on no risk list?
615 servers declare no tools we can read, because they build the list at runtime. They cannot be called read-only, so they appear on none of the risk hubs.

Keep going in your AI

Opens with this page as the source. We add nothing else to the prompt.