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/110

Verdicts · 3,145 servers

✓SAFE1,632!CAUTION521✕BLOCK992

Transports

A server can list more than one transport.

3,145 servers

2,617–2,640 of 3,145 · ranked by trust score, then stars

agentic-coding-harness

sblattj/agentic-coding-harness

✕BLOCK D

Run, watch, and meter Claude Code, Codex CLI, Gemini CLI, OpenCode, and Kiro from one CLI. Normalized events, cache-aware token usage, cost accounting verified against each agent's own records, budget caps, OpenTelemetry gen_ai spans, Langfuse traces, and a live web dashboard + MCP server.

★ 16 writestreamable-http

Blender Model Context Protocol

ahujasid/blender-mcp

✕BLOCK D

Community plugin to control Blender 3D with any LLM of your choice

★ 29,0361 write

Reverse Engineer Anything

morluto/rea

✕BLOCK D

Reverse engineer anything with agents, from app behavior down to native binaries.

★ 8,4701 writestdio

Iai

codeabra/iai-mcp

!CAUTION D

A cyber brain for your AI. It never forgets a detail, remembers exactly what you said, and learns how you work over time. Free, local, works with Cursor, Claude Code, Codex, OpenClaw, Hermes and more. MIT.

★ 9012 writestdio

Avrotize

clemensv/avrotize

✕BLOCK D

Avrotize is a command-line tool for converting data structure definitions between different schema formats, using Apache Avro Schema as the integration schema model.

★ 1312 writestdio

USolver

sdiehl/usolver

✕BLOCK D

A model context protocol server for solving combinatorial optimization problems with logical and numerical constraints.

★ 657 read

Keryx

actionhero/keryx

!CAUTION D

Keryx: The Fullstack TypeScript Framework for MCP and APIs

★ 354 readstreamable-http

Kindex

jmcentire/kindex

!CAUTION D

Knowledge index that learns from your conversations

★ 353 readstdio

lockstep

lockstep-team-agent/lockstep

✕BLOCK D

Keep your team's AI coding agents coordinated — without sharing source code.

★ 102 writestdio

memtomem-stm

memtomem/memtomem-stm

✕BLOCK D

Short-term memory proxy gateway with proactive memory surfacing for AI agents

★ 514 readsse

WebPieces TypeScript

deanhiller/webpieces-ts

✕BLOCK D
★ 42 read

obsidian-llm-wiki

Statutecontempt951/obsidian-llm-wiki

✕BLOCK D

Convert your Obsidian markdown vault into a six-persona MCP team for local LLM tools to provide cited, accurate insights.

★ 15 readstdio

assetfare-mcp

assetfare/assetfare-mcp

✕BLOCK D

USDC bridge API for AI agents and agent-wallet funding: Solana to Base + 76 routes; validated provider paths; exact 1bp; server never signs/submits.

★ 02 writestdio

vetagent

jakegu1/vetagent

✕BLOCK D

VetAgent — token risk intelligence MCP for AI agents. Cloudflare Worker deployment. Assess token risk (liquidity/honeypot/rug) and return actionable low/medium/high verdict. Remote MCP at https://vetagent.dev/mcp.

★ 01 readstreamable-http

ztrack

volter-ai/ztrack

✕BLOCK D

ztrack: evidence-gated task tracking for AI agents

★ 03 write

Memlab

facebook/memlab

✕BLOCK D

A framework for finding JavaScript memory leaks and analyzing heap snapshots

★ 5,06721 writestdio

Stripe Agent Toolkit

stripe/agent-toolkit

!CAUTION F

One-stop shop for building AI-powered products and businesses with Stripe.

★ 1,8331 writestdio

Solana Agent Kit

sendaifun/solana-agent-kit

!CAUTION F

connect any ai agents to solana protocols

★ 1,7138 writestdio

xmcp

basementstudio/xmcp

✕BLOCK D

The TypeScript MCP framework

★ 1,3373 writestreamable-http

sandbase-harness

sandbaseai/sandbase-harness

✕BLOCK D

Local-first, self-hosted AI agent runtime and MCP bridge with sandboxed sessions, memory, credentials, audit/replay, and a local Console.

★ 6473 writestdio

Kuvasz

kuvasz-uptime/kuvasz

✕BLOCK F

Kuvasz (pronounce as [ˈkuvɒs]) is an open-source uptime and SSL monitoring service, with multiple notification channels, status pages, IAC support via YAML, Prometheus integration, a complete REST API and many more!

★ 6361 read

Salesforce

salesforcecli/mcp

✕BLOCK D

MCP Server for interacting with Salesforce instances

★ 48315 readstdio

premiere-pro-mcp

leancoderkavy/premiere-pro-mcp

✕BLOCK D

Adobe Premiere Pro MCP — independent, local-first server for supported workflows. Connect Claude, Codex, or Cursor through a local CEP bridge.

★ 3252 writestdio

Remote PostgreSQL Database

coleam00/remote-mcp-server-with-auth

✕BLOCK D

Template for a remote MCP server with GitHub OAuth - following best practices for building MCP servers so you can take this as a starting point for any MCP server you want to build!

★ 3002 write

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.