MCP servers · least privilege

MCP servers with no destructive tools

3,151 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. 1,622 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 30 Oct 2026. https://modelsatlas.com/mcp/no-destructive/page/87

Verdicts · 3,151 servers

✓SAFE1,634!CAUTION522✕BLOCK995

Transports

A server can list more than one transport.

3,151 servers · 2,065–2,088 · ranked by trust score, then stars
pgtriage!CAUTION C

pgtriage/pgtriage

Read-only MCP server for evidence-based PostgreSQL performance auditing

★ 0GitHub stars1 writestdio
pi-mcp-adapter!CAUTION C

nicobailon/pi-mcp-adapter

Token-efficient MCP adapter for Pi coding agent

★ 1,588GitHub stars3 writestdio
Better Email!CAUTION C

n24q02m/better-email-mcp

IMAP/SMTP email for AI agents -- read, send, organize folders, and manage attachments across multiple accounts, with auto-discovery.

★ 33GitHub stars1 writestdio
OpenOSINT!CAUTION C

openosint/openosint

AI-powered OSINT agent with interactive REPL, MCP server, and CLI. 20 tools. Works with Claude, GPT-4, or local models. For authorized security research only.

★ 1,631GitHub stars10 readstdio
YantrikDB!CAUTION C

yantrikos/yantrikdb-server

Cognitive memory database for AI agents — consolidates duplicates, detects contradictions, fades stale memories via temporal decay. Rust, Apache-2.0, ships as library / MCP server / HTTP cluster.

★ 175GitHub stars3 writestdio
Linear!CAUTION C

tacticlaunch/mcp-linear

MCP server that enables AI assistants to interact with Linear project management system through natural language, allowing users to retrieve, create, and update issues, projects, and teams.

★ 147GitHub stars2 writestdio
OpenZIM!CAUTION C

cameronrye/openzim-mcp

OpenZIM MCP is a modern, secure, and high-performance MCP (Model Context Protocol) server that enables AI models to access and search ZIM format knowledge bases offline.

★ 144GitHub stars4 readsse
Code Indexer!CAUTION C

zxfgds/mcp-code-indexer

★ 34GitHub stars8 readstdio
ADK Redis!CAUTION C

redis-developer/adk-redis

Redis integration for Google Agent Development Kit (ADK) - Memory, Sessions, Search Tools, MCP

★ 30GitHub stars2 readstreamable-http
ymm4MCP!CAUTION C

SCPgamerscp/ymm4MCP

ゆっくりムービーメーカー4のMCPです。将来的にはゆっくり動画を全自動で作れるようにしたいです。

★ 26GitHub stars1 readstdio
Thally!CAUTION C

thallylabs/thally

Open-source docs platform and product knowledge layer that turns merged product changes into evidence-backed updates for human review.

★ 5GitHub stars9 writestdio
dgui-hypermem!CAUTION C

ctaxnagomi/dgui-hypermem

DGUI-HyperMem (DeckerGUI HyperMemory) - self-hosted hybrid memory MCP server on Cloudflare Workers with a JEV (Choice/Noul/Score) reasoning layer and a HuggingFace training-brain flywheel.

★ 2GitHub stars3 writestreamable-http
seo!CAUTION C

110kc3/seo

Open registry and discovery API for AI products, agent tools, and MCP services.

★ 1GitHub stars5 readstreamable-http
figloo!CAUTION C

g761007/figloo

Chrome extension + local MCP server that lets coding agents like Claude Code and Codex read Figma designs through the web UI with view access: walk layers, inspect properties, take screenshots, export assets. No Figma API; never edits the design.

★ 0GitHub stars15 readstdio
Brasil!CAUTION C

jxnxts/mcp-brasil

MCP Server para 70 APIs públicas brasileiras

★ 1,786GitHub stars10 readstreamable-http
Brasil!CAUTION C

mcp-brasil/mcp-brasil

MCP Server para 70 APIs públicas brasileiras

★ 1,783GitHub stars10 readstreamable-http
intuition-lab-persome!CAUTION C

intuition-lab/personal-model

Build your HUMAN.md.

★ 1,333GitHub stars2 writesse
Blockrun Mcp!CAUTION C

BlockRunAI/blockrun-mcp

Live data for AI agents — search, research, markets, crypto, X/Twitter. Pay-per-call via x402 micropayments.

★ 391GitHub stars19 readstdio
Perplexity!CAUTION C

automations-project/vscode-perplexity-mcp

Perplexity AI search, reasoning, research, and compute - MCP server, dashboard, and multi-IDE auto-config for VS Code.

★ 77GitHub stars17 readstdio
Slack!CAUTION C

zencoderai/slack-mcp-server

★ 75GitHub stars2 writestdio
Nexus!CAUTION C

adawalli/nexus

MCP Server to make searching openrouter easy

★ 23GitHub stars3 readstdio
browser-gateway!CAUTION C

browser-gateway/browser-gateway

Control plane for browser fleets. One endpoint, any provider: failover, persistent profiles, session replay, REST API, MCP server, dashboard. OpenRouter for cloud browsers.

★ 18GitHub stars3 writestdio
ask-marcel-office-cli!CAUTION C

ask-marcel/ask-marcel-office-cli

The Microsoft 365 command line built for AI agents: one sign-in, 182 read-only commands, every document as clean markdown. Outlook, OneDrive, SharePoint, Teams, Calendar, Excel, OneNote, Planner. MCP server included.

★ 9GitHub stars2 writestdio
shyn!CAUTION C

shyn-labs/shyn

shyn — a register of your days. Local-first memory for your Mac: pages, meetings, notes — encrypted on-device, recalled by your AI over MCP.

★ 1GitHub stars6 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 (1,622 servers). This list also includes servers whose tools write: 1,529 of the 3,151 do.
Does no destructive tools mean safe?
No. 1,634 are SAFE, 522 CAUTION and 995 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.