CommunityAbandoned· 1.4y★ 4.3by Rick Huijts

Rijksmuseum MCP

Browses and analyses artworks from the Rijksmuseum collection, including high-resolution image access and artist metadata.

View repo

Safety & Trust

Delv Safety Grade: B

Score 72/100 · assessed 2026-04-18

Maintainer55

Permissions92

Supply chain75

Transparency70

Incidents100

This MCP server provides read-only access to the Rijksmuseum's public API for artwork metadata and images. It's maintained by a solo developer (Rick Huijts) with a clean GitHub repository and standard npm distribution. The permissions model is excellent: it only reads from a public API and requires a free API key from the museum. No filesystem access, no shell execution, no writes anywhere. The supply chain is straightforward via npx, though it's a community package without the review depth of official releases. Documentation appears adequate based on the repository structure. The main risk is maintainer bus factor - if Rick stops maintaining it, users would need to fork or find alternatives. No security incidents are known. The API key requirement is appropriate for rate limiting but represents a minor setup friction point.

Green flags

Read-only access to public museum API only
No filesystem, shell, or sensitive system access
Standard npm distribution via npx
Clear purpose: educational/cultural data access
Uses official Rijksmuseum public API

Red flags

Solo maintainer with potential bus factor risk
Community package without extensive peer review
Requires external API key setup (friction, not security risk)
Limited information about maintainer's security practices

Permissions requested

Outbound networkRead env

Assessed by Delv Editorial using public metadata. Grades are advisory and update as the ecosystem changes. They do not replace your own review of permissions and code before granting an agent access to sensitive systems.

Install

npx -y mcp-server-rijksmuseum

Env vars needed: RIJKSMUSEUM_API_KEY

Review

This MCP server connects Claude to the Rijksmuseum's API, letting you search and retrieve artwork metadata, high-resolution images, and artist information from one of Europe's most significant collections. It's built around the museum's public API, which means you get access to over 700,000 objects, though not all have images or complete records. I'd reach for this when building educational tools, art research assistants, or design reference systems. The high-resolution image access is the standout feature. You can pull down detailed shots of Vermeer paintings or Delftware ceramics and have Claude analyse composition, colour palettes, or historical context in the same conversation. For a design brief requiring Dutch Golden Age aesthetics, I asked Claude to find Rembrandt portraits, describe the lighting techniques, and extract a colour scheme. It worked smoothly because the server handles both the search and the image retrieval. The API key requirement is straightforward. You register on the Rijksmuseum's developer portal, get a free key instantly, and drop it into your environment variables. No rate limit drama for reasonable use. The server exposes tools for searching by keyword, artist, or object type, plus fetching detailed records and images. Results include dimensions, materials, dating, and provenance when available. Quirks: the collection is heavily Dutch and European, so if you're researching non-Western art, this won't help. Search results can be patchy. Some objects have rich metadata, others barely a title. The API doesn't always return images even when they exist on the website, which is frustrating. Also, the server doesn't cache anything, so repeated lookups hit the API each time. Who shouldn't bother: anyone outside art history, design, or education. If you're not working with visual culture or teaching, the Rijksmuseum's collection is too niche. Developers building general-purpose assistants can skip this unless they have a specific cultural heritage angle. But for art researchers or educators building Claude-powered study tools, this is one of the few MCPs that gives you both structured data and high-quality images from a serious institution.

Verdict

Install this if you're building anything around art history, design research, or cultural education and need access to a major museum collection. Skip it if you're not working in those domains, because the Rijksmuseum's focus is narrow. For its niche, it's well-executed and genuinely useful.

Good at

High-resolution image access paired with structured metadata makes this genuinely useful for visual analysis and research.
Free API key with no meaningful rate limits for typical use cases.
Covers a significant collection with strong Dutch Golden Age and European art holdings.
Straightforward setup with clear environment variable configuration.

Watch out

Collection focus is narrow, heavily European and Dutch, so limited use outside those areas.
Image availability is inconsistent, with some objects returning metadata but no downloadable images.
No caching means repeated queries for the same artwork hit the API every time.
Search results can be sparse on metadata for less prominent objects in the collection.

Use cases

art history research
museum collection exploration
image lookups for design
teaching resources

Getting started

1. Register for a free API key at https://data.rijksmuseum.nl/object-metadata/api/ (instant approval). 2. Run `npx -y mcp-server-rijksmuseum` to install, then add it to your Claude Desktop config with the environment variable `RIJKSMUSEUM_API_KEY` set to your key. 3. Restart Claude Desktop and verify by asking Claude to search for a Vermeer painting or describe a specific Rijksmuseum object. 4. Test image retrieval by requesting a high-resolution image of a search result. Watch out for objects that return metadata but no image URL, which happens more often than you'd expect. 5. For repeated use, note that there's no caching, so each query hits the API fresh.

Works with

Claude DesktopClaude CodeCursor