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/62

Verdicts · 3,151 servers

✓SAFE1,634!CAUTION522✕BLOCK995

Transports

A server can list more than one transport.

3,151 servers · 1,465–1,488 · ranked by trust score, then stars
mcp-server✓SAFE B

negm17111995/mcp-server

Remote MCP server for hotel and flight search on MAQAMI (3M+ hotels). Search live rates, read hotel details, then send the customer a secure checkout link on book.maqami.co. Streamable HTTP, no API key.

★ 2GitHub stars1 readstdio
MUSTer_MCP✓SAFE B

ABINETHACKER/MUSTer_MCP

📚 Simplify LLM access to the M.U.S.T. campus system, enabling smooth interactions with schedules, assignments, and course materials.

★ 2GitHub stars1 writestdio
redfox-mcp✓SAFE B

redfox-data/redfox-mcp

MCP server for RedFoxHub — 40 ready-to-use tools for Douyin, Xiaohongshu, WeChat, Bilibili, Toutiao & TikTok data, plus AI search (Kimi/Doubao/DeepSeek) and AI image/video generation. Works with dsh, Claude Code, Cursor and any MCP client.

★ 2GitHub stars3 writestreamable-http
trainingpeaks-mcp!CAUTION B

haytam111234/trainingpeaks-mcp

🏋️♂️ Connect TrainingPeaks to AI assistants with ease. Query workouts, analyze data, and track fitness trends without API approval hassles.

★ 2GitHub stars1 writestdio
Tin Can✓SAFE B

BrutalSystems/tincan

Peer messaging between live Claude Code, Codex and opencode sessions on one machine. One binary, run as an MCP server inside each.

★ 1GitHub stars2 writestdio
DuckDuckGo MCP Server✓SAFE B

ericthered926/duckduckgo-mcp-server

an mcp server to access DuckDuckGo

★ 1GitHub stars2 readstdio
AVCS — Agentic Version Control System✓SAFE B

izagood/avcs

Agentic Version Control System — an AI-native VCS where intent, session, operation, evidence and decision are first-class objects and code is a materialized projection of an operation graph.

★ 1GitHub stars12 writestdio
adsmart-app✓SAFE B

ZenniTTy/adsmart-app

Gerencie suas campanhas do Google Ads conversando com o Claude. Extensão open source para o Claude Desktop (MCPB), 100% local, com prévia, confirmação e desfazer.

★ 1GitHub stars8 read
agentRamen✓SAFE B

palrajjp/agentRamen

Local-first codebase intelligence memory and task context for coding agents, via CLI, MCP, and local HTTP.

★ 1GitHub stars10 readstdio
agentskills-mcp✓SAFE B

Drhir2460/agentskills-mcp

Search, read, and download curated agent skills via MCP for fast discovery and local use in agent workflows

★ 1GitHub stars2 write
askjev✓SAFE B

pZacca/askjev

Unofficial MCP server for Jev (Typesafe AI)

★ 1GitHub stars1 readstdio
baibai✓SAFE B

Zeon7744/baibai

MCP toolkit for Vibe Coding - 12 tools (format checker, data analyzer, README generator, TTS) with GitHub Releases distribution

★ 1GitHub stars1 writestreamable-http
byte-mcp-server✓SAFE B

0rkz/byte-mcp-server

MCP server for PayPerByte — pay-per-call USDC data feeds for AI agents (x402 on Base; EIP-712 attestations). No API keys, no token.

★ 1GitHub stars1 writestdio
capitoline✓SAFE B

marco-fabbri/capitoline

Self-hosted OpenAI-compatible gateway and MCP server over the official Claude Code, Codex and Antigravity CLIs, with a council of models

★ 1GitHub stars4 readstreamable-http
chrischall-hemnet✓SAFE B

chrischall/hemnet-mcp

Hemnet.se real estate for Claude: search, sold prices, listings & market stats

★ 1GitHub stars10 readstdio
cloudscape-docs-mcp!CAUTION B

prem676/cloudscape-docs-mcp

🌐 Enable seamless semantic search over AWS Cloudscape documentation for AI agents and coding assistants with this efficient MCP server.

★ 1GitHub stars2 read
context!CAUTION B

timmo001/context

🧭 Standalone CLI and MCP server for deterministic repository context

★ 1GitHub stars3 write
crawlora-mcp✓SAFE B

Crawlora-org/crawlora-mcp

Hosted MCP server for structured public web data: 3,631 tools across 471 platform groups (search, maps, e-commerce, social, finance, app stores), returning clean JSON. Works with Claude, Cursor, VS Code and any MCP client.

★ 1GitHub stars3 readstdio
danielortiz0220-nyc-restrooms✓SAFE B

danielortiz0220/nyc-restroom-mcp

Find NYC public restrooms and their latest Parks inspection status, from NYC Open Data

★ 1GitHub stars2 readstdio
Design-components-for-AI-agents!CAUTION B

richawo/Design-components-for-AI-agents

Animated React, Tailwind CSS and React Native components for AI coding agents. Source code, design prompts and JSON briefs, with a shadcn registry, CLI and MCP for Claude Code, Cursor and Codex.

★ 1GitHub stars2 writestdio
faceit-mcp✓SAFE B

saivi3171/faceit-mcp

Connect Claude to live CS2 FACEIT data to retrieve player stats, match history, ELO trends, and head-to-head comparisons via the Model Context Protocol.

★ 1GitHub stars8 read
flowfig✓SAFE B

iamalvisng/flowfig

Your coding agent draws animated diagrams from your code, and flowfig checks each box and arrow against that code. CI fails when a diagram drifts from the code. The output is one animated SVG with no script, for a README, PR or issue. Works with Claude Code, Cursor, GitHub Copilot, Codex, Gemini CLI

★ 1GitHub stars1 write
forcedream✓SAFE B

forcedreamai/forcedream-mcp

MCP server for ForceDream — discover, invoke, and trustlessly verify AI agents with cryptographic proofs.

★ 1GitHub stars1 writestdio
gfc✓SAFE B

jjjorfi/gfc

git-for-context — LLM context dedup: unchanged file re-reads collapse to hash-verified stubs, changed files return diffs. OpenCode plugin + zero-dep MCP server. Measured −34% per-turn context.

★ 1GitHub stars1 read

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.