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/36

Verdicts · 1,620 servers

✓SAFE913!CAUTION261✕BLOCK446

Transports

A server can list more than one transport.

1,620 servers

841–864 of 1,620 · ranked by trust score, then stars

trace.moe-mcp

soruly/trace.moe-mcp

✓SAFE B

Model Context Protocol (MCP) server for trace.moe anime scene search

★ 15 readstdio

web

GuionAI/web

✓SAFE B

A web research toolkit with multi-provider search and clean Markdown extraction from static and JavaScript-rendered pages, plus public code and library docs search—available via CLI and MCP.

★ 114 readstdio

web-search-mcp

awslew/web-search-mcp

✓SAFE B

编码 agent 查资料缺搜索 API key?免 key 搜多个引擎、优先可靠来源并抽取可读正文。 Search the web and read pages without an API key.

★ 13 readstdio

Fibel

k2b-dev/fibel

✓SAFE B

Docs first - everything else later

★ 03 readstreamable-http

@orboto/mcp

orboto/orboto-mcp

✓SAFE B
★ 02 readstdio

MCP TypeScript SDK !

redocly-contrib/mcp-typescript-sdk

✓SAFE B

The official TypeScript SDK for Model Context Protocol servers and clients

★ 02 readsse

aemet-client

Alysiadeceptive64/aemet-client

✓SAFE B

Access Spanish weather data through a TypeScript SDK and Model Context Protocol server for the AEMET OpenData API.

★ 03 readstdio

aeo-geo-mcp

SEOSiri-Official/aeo-geo-mcp

✓SAFE B

A sovereign, local-first MCP server for AI Engine Optimization (AEO), Generative Engine Optimization (GEO), llm.txt auditing, and content stickiness analytics.

★ 010 readstdio

agentsafepath

saviodamato/agentrank-mcp

✓SAFE B

Find real, verifiable businesses with provenance and source URLs - for AI agents.

★ 02 readstdio

AIMentionTracker

aimentiontracker/mcp

✓SAFE B

MCP server for AIMentionTracker — let Claude, Cursor and other MCP clients read how often ChatGPT, Claude, Gemini, Perplexity and Google AI Overviews name and cite your brand

★ 06 readstdio

aioproductos

aioproductos/claude-plugin

✓SAFE B

Official AIOProductOS plugin for Claude Code — connect your workspace (39 MCP tools) and run evidence-backed weekly product triage

★ 04 readstdio

airlock-rls-mcp

mateuszingano/airlock-rls-mcp

✓SAFE B

MCP server: let Cursor / Claude Code check a Supabase app for an exposed service_role key and broken RLS while it builds it

★ 02 read

biopharma-mcp

SEOSiri-Official/biopharma-mcp

✓SAFE B

An open-source, local-first Model Context Protocol (MCP) server written in TypeScript for **Biopharma Software Infrastructure, FDA 21 CFR Part 11 GxP Audit Trailing, CDISC SDTM/Allotrope Data Exports, 4PL Dose-Response Curve Fitting, and HIPAA PII/PHI Redaction**.

★ 010 readsse

biorobotics

SEOSiri-Official/biorobotics

✓SAFE B

An open-source, stateless Model Context Protocol (MCP) framework bridging biological datasets (NCBI/UniProt) with deterministic physical coordinates for robotic automation.

★ 06 readsse

bitmacrocode-charm

bitmacrocode/charm-mcp

✓SAFE B

MCP server: authoritative OEM automotive repair data (part numbers, torque specs, procedures) for 1982-2013 vehicles, from factory service manuals. Built for AI agents.

★ 02 readstdio

bubble-agent-mcp

0308ya9392/bubble-agent-mcp

✓SAFE B

Find emerging conversations before they blow up. An MCP server that brings live Bubble Agent trend intelligence to Claude, Codex, Cursor, and more.

★ 05 read

career-workspace-ambujsystems

Ambuj123-lab/career-workspace-ambujsystems

!CAUTION B

Evidence-Grounded Career Workspace & Claim-Validation Engine • Real-Time Company Intelligence • 0% Overclaim Guarantee

★ 08 readstdio

claude-conf

Vleickzs/claude-conf

✓SAFE B

Streamline Claude Code with a modular config toolkit for faster, cleaner terminal workflows

★ 03 read

codeparse

cmgoffena13/codeparse

!CAUTION B

Give agents fast, accurate, and compact context on your Python codebases

★ 08 readstdio

content-schema-mcp

SEOSiri-Official/content-schema-mcp

✓SAFE B

An open-source, local-first (MCP) server for Automated Schema org JSON-LD Generation, GA4 Metric Validation, llm.txt Auditing

★ 010 readstdio

crawler

mark1russell7/crawler

✓SAFE B

polite-crawler: a polite web crawler for research and AI agents. Fetch once, keep it, read it many times. robots.txt, one request per site, full-text search, MCP server.

★ 04 readstdio

Currensor

Kay0sstheory/Currensor

✓SAFE B

Live currency converter — a one-file web app, plus a free plugin for Claude and ChatGPT with an in-chat card. 166 currencies, no sign-in.

★ 01 readstreamable-http

dns-sec-audit-mcp

SEOSiri-Official/dns-sec-audit-mcp

✓SAFE B

A sovereign, local-first MCP server for Technical SEO DNS record auditing, SOA health checks, SSL/TLS inspection, and HTTP security header analysis.

★ 010 readstdio

etl-pipeline-mcp

SEOSiri-Official/etl-pipeline-mcp

!CAUTION B

A sovereign, local-first MCP-Native ETL System with Hot/Cold tiering, PII scrubbing, ID stitching, Parquet buffers, and data warehouse exports.

★ 09 readstdio

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.