protogrid

Search / space.r3plica/catalog

R3PLICA AI Connector

R0· no sign-in

Catalog of certified digital replicas from premium design brands, for any MCP-compatible AI agent.

v2.3.3 · active · repository · website · descriptor JSON · versions

An agent can connect right now, no human step.Remote endpoint, no authentication, reachable on the last probe.
Quality79good
Trust88of 100
Uptime, 30 days97%
Latency p50537 ms

Connect

8 clients · secrets stay placeholders
{
  "mcpServers": {
    "catalog": {
      "type": "http",
      "url": "https://mcp.r3plica.space/mcp"
    }
  }
}
claude mcp add --transport http catalog https://mcp.r3plica.space/mcp
[mcp_servers.catalog]
url = "https://mcp.r3plica.space/mcp"
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "catalog": {
      "type": "remote",
      "url": "https://mcp.r3plica.space/mcp",
      "enabled": true
    }
  }
}
{
  "mcpServers": {
    "catalog": {
      "type": "http",
      "url": "https://mcp.r3plica.space/mcp"
    }
  },
  "deeplink": "cursor://anysphere.cursor-deeplink/mcp/install?name=catalog&config=eyJ0eXBlIjoiaHR0cCIsInVybCI6Imh0dHBzOi8vbWNwLnIzcGxpY2Euc3BhY2UvbWNwIn0%3D"
}
{
  "servers": {
    "catalog": {
      "type": "http",
      "url": "https://mcp.r3plica.space/mcp"
    }
  }
}
{
  "mcpServers": {
    "catalog": {
      "httpUrl": "https://mcp.r3plica.space/mcp"
    }
  }
}
extensions:
  "catalog":
    enabled: true
    name: "catalog"
    type: streamable_http
    uri: "https://mcp.r3plica.space/mcp"
    timeout: 300

Quality 79 / 100good

protocol75 · weight 25
  • warn
    protocol.modernnewest supported version is 2025-06-18; 2026-07-28 not supported, older versions are deprecated until 2027-07-28
  • n/a
    protocol.statelessonly applies to 2026-07-28 servers
  • pass
    protocol.transportstreamable HTTP
  • n/a
    protocol.list_ttlonly applies to 2026-07-28 servers
authorizationn/a · weight 15
  • n/a
    auth.prm, auth.as_metadata, auth.cimdno remote uses OAuth
  • n/a
    auth.secret_in_urlno templated URL
tool hygiene79 · weight 30
  • pass
    tools.descriptionsevery tool has a description
  • warn
    tools.description_length2 tools have descriptions over 1,024 characters, which crowds the context
  • pass
    tools.schemasevery tool has an object input schema
  • pass
    tools.annotations16 of 16 tools declare readOnlyHint or destructiveHint
  • fail
    tools.directory_hintsno tool declares all four of readOnlyHint, destructiveHint, idempotentHint and openWorldHint; missing: brand_get (destructiveHint, idempotentHint, openWorldHint); product_collection_list (destructiveHint, idempotentHint, openWorldHint); product_search (destructiveHint, idempotentHint, openWorldHint); …
  • pass
    tools.token_costabout 2,610 tokens to load every tool
  • pass
    tools.api_dumptools are not a one-to-one API dump
stability100 · weight 15
  • pass
    stability.changesno tool changes in 30 days
  • pass
    stability.rug_pullno tool changed its meaning under the same name
dependencies67 · weight 15
  • pass
    deps.known_vulnsno known advisory in 99 resolved packages
  • fail
    deps.mcp_sdk_version@modelcontextprotocol/sdk 1.30.1 has 1 known advisory (GHSA-6qxp-vccf-f47h); update @modelcontextprotocol/sdk to 1.31.0 or later
  • pass
    deps.resolvable[email protected] resolved: 99 packages

16 tools, about 2,610 tokens to load them all · tool set 4592062e27c5 · checked 2026-10-20 17:34 UTC

Tool changes

No tool definition changes recorded. A server's first observation is its baseline; later probes record what changes.

Checks run on what our credential-free, read-only probes observe; tools are never called and no code audit is performed. How it is computed.

Trust 88 / 100

provenance85
liveness99
freshness65
hygiene100

+repository +dns-namespace +website +reachable +uptime-97% ~updated-130d-ago

Derived from observable signals (official registry feed and our own credential-free probes); no code audit performed. How it is computed.

Dependencies

[email protected] npm98 dependencies, resolved · MCP SDK @modelcontextprotocol/sdk 1.30.1 · advisories checked 2026-10-19

advisoryseveritypackagefixed in
GHSA-6qxp-vccf-f47hMCP TypeScript SDK: OAuth client could send credentials to an authorization server chosen by the MCP serverhigh@modelcontextprotocol/[email protected]MCP SDK, directfixed in 1.31.0

Dependency graphs from deps.dev; advisories from OSV.dev, including the GitHub Advisory Database and the PyPI Advisory Database; all CC BY 4.0. The graph is what a clean install of this version resolves today; nothing was installed or run. How it is checked.

Endpoints

remoteauthreachableuptime 30dp50protocollast ok
https://mcp.r3plica.space/mcp streamable-httpnoneyes97%537 ms2025-06-182026-10-20
https://mcp.r3plica.space/sse ssenoneyes96%535 ms2025-06-182026-10-20
packageregistryversionruntimesecrets
r3plicanpm2.3.3–

Tools (16)

brand_getread-onlyDetailed information about a specific brand by name (description, country, designers, catalog presence). Triggers: 'tell me about <brand>', 'info on <brand>', 'who is <brand>', 'parlami di <brand>', 'chi è <brand>'.
brand_listread-onlyList all R3PLICA partner brands (default limit 200). Triggers: 'what brands do you have', 'list the brands', 'which brands are on R3PLICA', 'quali brand avete', 'elenca i brand', 'che marche ci sono'.
brand_searchread-onlySearch R3PLICA brands by name substring. Use when the user mentions a brand but you need to resolve the exact catalog name. Triggers: 'find the brand <x>', 'is <x> on R3PLICA', 'cerca il brand <x>', 'avete <x>'.
finish_collection_getread-onlyGet a finish collection detail with all its individual finishes. Response includes `catalog_url` (deep-link to the collection page on catalog.r3plica.space), `finish_directory[]` (compact flat list with name + mongo_id + asset_name + thumbnail_url — designed to survive context truncation), and `finish_details[]` (full nested data). CRITICAL when generating documents/artifacts: copy the `thumbnail_url` and `catalog_url` fields VERBATIM. Do NOT construct catalog URLs yourself — the route is `/go/finish/<mongo_id>` (singular 'finish', NOT 'finish-collection' or 'finish/collection'). Do NOT construct thumbnail URLs from asset_name. If you need data for a single finish whose parent collection has been pushed out of context, call `finish_get` instead.
finish_collection_listread-onlyList R3PLICA finish collections (e.g. fabric, leather, wood, marble, metal sets sold as one asset). Optional brand/category filters. Default limit 5 collections per call — each can contain 30-50 individual finishes once enriched, so keep limits modest. TERMINOLOGY: a `finish_collection` is the asset sold on the platform; it contains many individual `finishes` inside `finish_details[]`. Do NOT call individual finishes 'variants' — 'variant' is reserved for product dimensional/additional variants. Response shape: { _meta: { terminology, asset_paths, deep_link_routes, thumbnail_rule, catalog_url_rule }, data: [collections] }. Each collection has a `catalog_url` and a flat `finish_directory[]`; each finish in `finish_details[]` carries a `thumbnail_url`. CRITICAL when embedding images: use the `thumbnail_url` field VERBATIM. NEVER construct CDN URLs from `asset_name` yourself — the file prefix is `MI_` (Material Instance) for ALL finishes regardless of brand; agents that infer brand-derived prefixes like `MO_`/`BO_` produce broken links.
finish_collection_searchread-onlySearch finish collections by name. Default limit 10. Response includes `_meta.terminology` + `_meta.asset_paths` notes and each finish in `finish_details[]` carries a `thumbnail_url`. CRITICAL: never construct CDN URLs from asset_name yourself — use the `thumbnail_url` field as-is.
finish_getread-onlyLook up a single individual finish by its mongo_id. Returns its exact `name` (e.g. 'Saffron', 'Walnut'), `asset_name`, `thumbnail_url`, `catalog_url` (deep-link to the PARENT collection page — individual finishes don't have their own page), and parent finish collection identity. Use this when composing documents and you need the precise URL for a specific finish without re-fetching the whole collection — avoids the URL-guessing trap when the collection has already been pushed out of your context window.
finish_searchread-onlySearch individual finishes across all collections by finish name (e.g. 'Walnut', 'Saffron') or asset_name substring. Optional `collection` filter narrows by finish collection name. Default limit 50. Returns each match with `asset_name`, `thumbnail_url`, and parent collection identity attached. Use when you know the finish name but not its mongo_id — e.g. 'find me the Walnut finish in the Bolzan Wood collection'.
product_collection_listread-onlyList product collections (a brand's product lines, e.g. 'Eclipse', 'Charisma'). Default limit 100. A product collection is a named product line within a brand — NOT a finish set (use finish_collection_* for those). Triggers: 'what collections does <brand> have', 'list <brand> product lines', 'quali collezioni ha <brand>', 'le linee di <brand>'.
product_collection_searchread-onlySearch product collections by name substring (product lines, not finish sets). Triggers: 'find the <x> collection', 'cerca la collezione <x>', 'la linea <x>'.
product_compareread-onlyCompare two products side by side. Returns two markdown cards under `## Product 1` / `## Product 2`, each with a 2-column spec table. Triggers: 'compare X and Y', 'X vs Y', 'which is bigger/better', 'differenza tra X e Y', 'confronta X e Y', 'meglio X o Y'. CRITICAL: preserve both card structures verbatim — the side-by-side tables are how the user visually compares.
product_getread-onlyGet detailed information about a specific product. Returns one structured markdown card (heading + `![](thumbnail_url)` image + spec table + description + catalog link). CRITICAL: preserve the markdown image and table verbatim. Do not flatten into prose. Do NOT reconstruct the thumbnail URL from asset_name — copy the provided `![](url)` line as-is.
product_listread-onlyList R3PLICA products with optional structured filters (brand, category, collection, designer, year). Use this for browsing/filtering by attribute (e.g. 'all Cassina sofas', 'chairs from 2020'); use product_search instead when the user gives a free-text query or keyword. Triggers: 'show me <category>', 'list <brand> products', 'browse the catalog', 'mostrami <categoria>', 'fammi vedere i prodotti di <brand>', 'sfoglia il catalogo'. Default limit is 50 products per call — raise via the `limit` argument if needed. Returns a sequence of structured markdown product cards: each has a `###` heading with name and brand, an inline `![](thumbnail_url)` markdown image, a 2-column markdown table with the spec rows (Designer, Year, Collection, Dimensions, Materials, Category), a blockquote description, and a catalog deep-link. CRITICAL when replying to the user: emit the cards VERBATIM as returned. Keep every `###` heading, the full markdown image line, the markdown table (`| ... |`), the blockquote, and the `[View on R3PLICA catalog →](url)` link untouched. Do NOT flatten the table into a comma-separated paragraph. Do NOT rewrite specs as prose. Do NOT construct alternate URLs from the asset_name — the CDN prefix is `SM_` (Static Mesh) for all products regardless of brand; pattern-matching brand-derived prefixes produces broken links.
product_searchread-onlySearch R3PLICA products by free-text query. Returns structured markdown product cards (heading + thumbnail image + spec table + description + catalog link). Use this when the user gives a keyword or descriptive phrase (e.g. 'find a curved velvet sofa', 'lounge chairs'); use product_list when filtering purely by structured attributes. Triggers: 'find <x>', 'search for <x>', 'do you have <x>', 'I'm looking for <x>', 'cerca <x>', 'trova <x>', 'avete <x>', 'sto cercando <x>'. Default limit is 50. CRITICAL: preserve the markdown structure verbatim — keep the `![](url)` image lines and the tables, do not flatten into prose. NEVER reconstruct image URLs from asset_name yourself — the CDN file prefix is constant (`SM_`), not brand-derived.
product_statsread-onlyCatalog statistics grouped by category, brand, designer or year. Returns aggregate counts only — small, fixed-size payload. Triggers: 'how many products', 'catalog breakdown', 'how many sofas/chairs', 'products per brand', 'quanti prodotti', 'quanti per brand', 'statistiche catalogo', 'quante sedie avete'.
style_guide_getread-onlyReturn the R3PLICA brand and output style guide as markdown. Covers card structure, typography, color palette, image aspect ratio (always square), hierarchy, mandatory link footer, and ready-to-copy empty card scaffolds (product card, finish swatch, both markdown and HTML). CALL THIS FIRST whenever you are about to generate any document, report, artifact or branded output about R3PLICA catalog content. The guide ensures outputs feel consistent across users and align with the platform's visual identity. Costs ~2KB of context but saves the conversation from inconsistent formatting and from inventing fields the platform doesn't expose (prices, lead times, certifications).

Schemas: list_tools.

Similar MCP servers

Official server.json
{
  "name": "space.r3plica/catalog",
  "title": "R3PLICA AI Connector",
  "$schema": "https://static.modelcontextprotocol.io/schemas/2025-12-11/server.schema.json",
  "remotes": [
    {
      "url": "https://mcp.r3plica.space/mcp",
      "type": "streamable-http"
    },
    {
      "url": "https://mcp.r3plica.space/sse",
      "type": "sse"
    }
  ],
  "version": "2.3.3",
  "packages": [
    {
      "version": "2.3.3",
      "transport": {
        "type": "stdio"
      },
      "identifier": "r3plica",
      "registryType": "npm"
    }
  ],
  "repository": {
    "url": "https://github.com/R3PLICAsrl/r3-cli",
    "source": "github"
  },
  "websiteUrl": "https://www.r3plica.space/integrations/mcp",
  "description": "Catalog of certified digital replicas from premium design brands, for any MCP-compatible AI agent."
}