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/39

Verdicts · 1,620 servers

✓SAFE913!CAUTION261✕BLOCK446

Transports

A server can list more than one transport.

1,620 servers

913–936 of 1,620 · ranked by trust score, then stars

mcp-openei-urdb

pipeworx-io/mcp-openei-urdb

✓SAFE B

OpenEI Utility Rate Database (URDB) MCP

★ 02 readstdio

mcp-pjm

pipeworx-io/mcp-pjm

✓SAFE B

Real-time electricity market data from PJM Interconnection, the largest US grid operator (13 states plus Washington DC, about 65 million people): wholesale power prices by node, electricity generation by fuel type, and the seven-day demand forecast.

★ 04 readstdio

mcp-planetary-computer

pipeworx-io/mcp-planetary-computer

✓SAFE B

Earth-observation imagery catalogue from Microsoft's Planetary Computer — search its open satellite and environmental collections (Sentinel, Landsat, NAIP, Daymet, ERA5, biodiversity) by area and date and get item ids, footprints, dates and asset URLs.

★ 02 readstdio

mcp-ptitchef

smeet666/mcp-ptitchef

✓SAFE B

MCP server for Ptitchef. Browse the recipe category tree and read its recipes. No API key required.

★ 07 readstdio

mcp-supertoinette

smeet666/mcp-supertoinette

✓SAFE B

MCP server for Supertoinette. Read French recipes and rescale their quantities honestly. No API key.

★ 06 readstdio

mcp-thai-text

anthonychkwn/mcp-thai-text

✓SAFE B

MCP server giving AI agents correct Thai text handling: word segmentation, caption line wrapping, and tone-mark-safe rendering

★ 03 read

mcp-zoom

pipeworx-io/mcp-zoom

✓SAFE B

Zoom MCP Pack

★ 04 readstdio

mibly-apple-health-mcp

littlenew117/mibly-apple-health-mcp

✓SAFE B

Connect Apple Health to any MCP-compatible AI agent — sleep, steps, heart rate, HRV, workouts and more, via the Mibly iPhone app.

★ 02 readstdio

millerchou-strava-bridge

millerchou/strava-mcp-bridge

!CAUTION B

Use Strava's official MCP from Codex and OpenCode without creating a developer app. Local macOS Keychain and privacy-first streams.

★ 03 readstdio

nara-catalog-mcp

ianderso/nara-catalog-mcp

✓SAFE B

MCP server for genealogical research in the US National Archives Catalog: search records, read OCR text and citizen transcriptions, browse the hierarchy, and download page images.

★ 02 readstdio

nightmarquee-mcp

ant-dev-lab/nightmarquee-mcp

✓SAFE B

Cinematic website prompts with live previews, inside your editor. MCP server for Claude Code, Claude Desktop and Cursor.

★ 05 readstdio

nofax

AKzar1el/nofax

✓SAFE B

Human-in-the-loop approvals and notifications for AI coding agents via CLI, MCP, Claude Code, Codex and Cloudflare Workers.

★ 07 readstdio

open-signals

Civops-Platform/open-signals

!CAUTION B

Public data signals with provenance and licence on every reading: source catalogue, fetchers, REST, MCP, GeoJSON layers

★ 09 readstreamable-http

pazair

Kula-Labs/pazair

✓SAFE B

Where agents keep their word. A marketplace where AI agents buy from and sell to other AI agents (MCP, pay on delivery, signed receipts).

★ 01 read

pullpush

kaduno-systems/pullpush-mcp

✓SAFE B

Connect any two APIs and keep them in sync — official MCP server for PullPush.ai. 45 tools: shadow-mode previews, canonical data models, diagnostics.

★ 01 readstdio

r3tr0eth-emap

r3tr0eth/emap-labs

!CAUTION B

emap/labs — infraestructura de inteligencia geográfica: datasets, evals y RAG geoespacial para Euskadi (ES/EU)

★ 05 readstreamable-http

rererr-amenbo

rererr/amenbo

✓SAFE B

Skims the web without making waves — a Japanese-web-native MCP server for low-impact, token-efficient web collection

★ 03 readstdio

rutter

shinytoyrobots/rutter

✓SAFE B

A record of what your AI sessions decided — and the receipts behind it. An MCP memory-of-use server over markdown notes: append-only, content-hashed, no model inside.

★ 04 readstdio

search-governance-mcp

SEOSiri-Official/search-governance-mcp

✓SAFE B

An open-source, local-first Model Context Protocol (MCP) server for **AI Search Crawler Governance, Brand Safety Guardrails, Canonical Integrity, and IndexNow Notifications

★ 010 readstdio

seedance-prompts-mcp

hanshs474/seedance-prompts-mcp

✓SAFE B

MCP server: 150 AI video prompts that were each rendered into a real video with Seedance 2 — search, inspect, audit and compose. Zero dependencies.

★ 05 readstdio

selflabbs-security-intel

datakoot/security-intel-mcp

✓SAFE B

Vulnerability-intelligence MCP server for AI agents: CVE lookups (NVD) and dependency-manifest audits (OSV). Keyless.

★ 04 readstreamable-http

semantic-entity-mcp

SEOSiri-Official/semantic-entity-mcp

✓SAFE B

An open-source, local-first Model Context Protocol (MCP) server for Named Entity Extraction, Wikidata Disambiguation, sameAs Schema Linking, and Knowledge Graph Triple Construction

★ 010 readstdio

seo-research-mcp

Farbod148/seo-research-mcp

✓SAFE B

🔍 Streamline your SEO research with free tools designed for AI-powered IDEs, enhancing analysis and compliance with third-party services.

★ 04 read

seob5285-alt-x402-facilitator-proxy

seob5285-alt/x402-facilitator-proxy

✓SAFE B

"An open, multi-tenant HTTP 402 infrastructure gateway and Model Context Protocol (MCP) server daemon built for autonomous AI agents. Enables programmatic 1% platform-fee split settlements in USDC on Base L2 with plug-and-play templates for Express.js, LangChain (TS), and CrewAI (Python)."

★ 01 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.