MCP servers · least privilege

MCP servers with no destructive tools

3,145 MCP servers declare at least one tool and none of them is classed destructive: nothing deletes, overwrites or runs a command through a declared tool. 1,620 of them only read; the rest may create or edit.

Ask an AI about this pageGChatGPTCClaudePPerplexityGeGeminiXGrok
More
Cite

Cite this page

ModelsAtlas. “MCP servers with no destructive tools”. Retrieved 29 Oct 2026. https://modelsatlas.com/mcp/no-destructive/page/32

Verdicts · 3,145 servers

✓SAFE1,632!CAUTION521✕BLOCK992

Transports

A server can list more than one transport.

3,145 servers

745–768 of 3,145 · ranked by trust score, then stars

Whale Tracker

kukapay/whale-tracker-mcp

✓SAFE B

A mcp server for tracking cryptocurrency whale transactions.

★ 562 read

Adb

srmorete/adb-mcp

✓SAFE B

An MCP (Model Context Protocol) server for interacting with Android devices through ADB in TypeScript.

★ 552 writestdio

Aws Bedrock Integration

davidshtian/mcp-on-aws-bedrock

✓SAFE B

A simple and clear example for implementation and understanding Anthropic MCP (on AWS Bedrock).

★ 551 readstreamable-http

Azure AI Agents

farzad528/mcp-server-azure-ai-agents

✓SAFE B

Model Context Protocol Servers for Azure AI Search

★ 555 read

Headless Gmail

baryhuang/mcp-headless-gmail

✓SAFE B

A MCP (Model Context Protocol) server that provides get, send Gmails without local credential or token setup.

★ 551 writestdio

Mcp-Flowise

matthewhand/mcp-flowise

✓SAFE B
★ 551 writestdio

Note Post

go-555/note-post-mcp

✓SAFE B
★ 552 writestdio

OpenAlex Research

oksure/openalex-research-mcp

✓SAFE B

MCP server for the OpenAlex API — search 240M+ scholarly works, analyze citations, track research trends, and map collaboration networks

★ 5526 readstdio

Powerpoint

ichigo3766/powerpoint-mcp

✓SAFE B

A MCP server project that creates PowerPoint presentations, forked from supercurses/powerpoint with additional features

★ 5510 writestdio

RunAPI

runapi-ai/mcp

!CAUTION B

RunAPI MCP server for model discovery, pricing lookup, media task creation, and balance checks.

★ 552 writestdio

SearXNG Scraper

pwilkin/mcp-searxng-public

✓SAFE B

An MCP server that queries public SearXNG instances, parsing HTML contents into a JSON result

★ 551 read

Thoughtful Claude

martinbowling/thoughtful-claude

✓SAFE B

MCP server that enhances Claude's reasoning capabilities by integrating DeepSeek R1's advanced reasoning engine 🤔

★ 551 read

WolframAlpha LLM

garoth/wolframalpha-llm-mcp

✓SAFE B

An MCP Server for WolframAlpha's LLM API, able to return structured knowledge & solve math

★ 553 readstdio

Axon Pro

talhamsajid/axon-pro

✓SAFE B

Enterprise-grade graph-powered code intelligence engine. Deep structural awareness for Laravel, PHP, Python, and JS. Powers AI agents with high-fidelity architectural context via MCP & Neo4j.

★ 541 writestdio

Notify

cactusinhand/mcp_server_notify

✓SAFE B

Send system notification when Agent task is done.

★ 541 writestdio

PlantUML

infobip/plantuml-mcp-server

✓SAFE B
★ 544 readstdio

RDF Explorer

emekaokoye/mcp-rdf-explorer

✓SAFE B

A Model Context Protocol (MCP) server that provides conversational interface for the exploration and analysis of RDF (Turtle) based Knowledge Graph in Local File mode or SPARQL Endpoint mode.

★ 544 write

Real Browser

ofershap/real-browser-mcp

✓SAFE B

MCP server + Chrome extension that gives AI agents control of your real browser with existing sessions and logins

★ 545 writestdio

Beanquery

vanto/beanquery-mcp

✓SAFE B

Beancount MCP Server is an experimental implementation that utilizes the Model Context Protocol (MCP) to enable AI assistants to query and analyze Beancount ledger files using Beancount Query Language (BQL) and the beanquery tool.

★ 532 write

Claude Code Delegator

fjgbue/claude-delegator-deepseek-mcp

✓SAFE B

MCP server: delegate heavy-token tasks from Claude Code to DeepSeek, Kimi, GLM, Qwen, Grok, or any OpenAI-compatible model. Mix providers per task, cost receipt on every call. Zero dependencies.

★ 533 read

Claude Delegator

12122j/claude-delegator-deepseek-mcp

✓SAFE B

MCP server: delegate heavy-token tasks from Claude Code to DeepSeek, Kimi, GLM, Qwen, Grok, or any OpenAI-compatible model. Mix providers per task, cost receipt on every call. Zero dependencies.

★ 533 read

LinkedIn

dishant27/linkedin-mcp-server

✓SAFE B

Model Context Protocol (MCP) server for LinkedIn API integration

★ 531 writestdio

ROADrecon

atomicchonk/roadrecon_mcp_server

✓SAFE B

Claude MCP server to perform analysis on ROADrecon data

★ 539 read

Apple Notes

henilcalagiya/mcp-apple-notes

✓SAFE B

An local mcp server for interacting with Apple Notes via LLM.

★ 524 readstdio

Other risk hubs

By job

Questions

What counts as destructive?
A tool that deletes, overwrites or runs commands. We class every tool the source declares as read, write or destructive; a server is listed here when it has tools and none is destructive.
How is this different from read-only?
Read-only servers have only read tools (1,620 servers). This list also includes servers whose tools write: 1,525 of the 3,145 do.
Does no destructive tools mean safe?
No. 1,632 are SAFE, 521 CAUTION and 992 BLOCK on this list. The verdict comes from the audit of the source, not from the tool list alone.

Keep going in your AI

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