Skip to content
NobleBlocks logo

NobleBlocks MCP server

Which MCP servers are NobleBlocks's own, from the official registry and from asking its own domain, whether they answer, and what they offer. A program checked each of these, nobody copied them from a product page.

read 2026-10-31

← Back to NobleBlocks

Whose server it is
  • Registry namecom.nobleblocks/nobleblocks-mcpverified in the MCP registry

The registry name sits under nobleblocks.com. The registry only lets a publisher use that namespace after proving control of the domain through DNS or HTTP.

When we asked itchecked 2026-10-31
  • Stateanswerschecked 2026-10-31
  • Tools11
  • Server calls itselfnobleblocks
  • Server version4.0.0
  • Protocol version2025-06-18

On 2026-10-31 the server answered initialize and tools/list without a sign-in. First answer in 571 ms.

Should an agent connect itNot proven either way yet
  • Its tools do every use case we checked: search millions of academic papers by topic; filter paper search results with classic search filters.
  • Checked on 2 of the 7 days it takes to judge reliability.

Once connected

  • For "search millions of academic papers by topic", the likely first call is search_papers.
  • For "filter paper search results with classic search filters", the likely first call is search_papers.

Said by rules from the measurements above. The likely first call is a model's pick from the tool list; read its description before relying on it.

Measured dailydaily checks: 2, the first on 2026-10-31
  • INITE MCP scoreafter a weeka score needs 7 daily checks: 2 so far
  • What an agent can do with it2 of 2 use cases
  • Use1 GitHub stars
  • Answeredmeasuring2 of the 7 daily checks it takes to print a share
  • Median first answer613 ms
  • Space in an agent's context≈ 2,393 tokensthe definitions of all its tools (11), which an agent carries on every turn
  • Tool definitionsA · 100 of 100
  • Typed results11 of 11 toolsdeclared outputSchema: results come typed, read once
  • Every tool passes all six checks.

What an agent can do with it

  • can do · search millions of academic papers by topic
  • can do · filter paper search results with classic search filters
  • not sure · start a literature review from smart search results

Each of the product's use cases, put to Jev against this server's tool list: can an agent do it with these tools?

Worth a look

None of the five warning signs in these definitions.

A program checks the server once a day: initialize and tools/list, no model. Tokens are the definitions' length at about four characters a token. The grade averages six checks of every tool's name, description, schema and annotations; it says how well the tools are described, not whether they work. The score weighs what was measured and nothing else: answering (25%), doing the product's job (25%), definitions (20%), context (10%), use (10%) and how readily a client connects (10%). A part that was not measured is left out, and a hidden instruction caps the score at 40. The state of vendors' MCP servers: method and numbers →

Tools it offers11 tools · read 2026-10-31
  • search_papers

    Search the NobleBlocks index of scholarly papers by keyword relevance, with filters. Each result has title, authors, year, journal, DOI, citation count, an abstract excerpt, a NobleBlocks link and `best_link` (free full text when known…

  • search_trials

    Search clinical trial registrations (ClinicalTrials.gov or WHO ICTRP) held in the NobleBlocks index. Each result carries its registry id (`trial_id`, e.g. NCT01234567). Use get_paper with that id for the full record.

  • get_paper

    Full metadata for one paper or trial: authors, journal, volume/issue/pages, identifiers (DOI, PMID, arXiv, OpenAlex, trial id, NobleID ARK when confirmed), retraction status, open access link, citation count, abstract and links…

  • get_citing_papers

    Papers in the NobleBlocks index that cite this paper, most cited first. Citation links are still being loaded, so an empty answer is reported as such, not as 'never cited'.

  • get_references

    Papers this paper cites (its reference list) that are in the NobleBlocks index, most cited first.

  • search_authors

    Find researchers by name. Returns name, ORCID iD, OpenAlex id, affiliation and reported counts. Use get_author with an ORCID iD for the profile and top papers.

  • get_author

    One researcher's public profile by ORCID iD: name, affiliation, topics, reported metrics, NobleID when recorded, and their most cited papers in the index.

  • get_citation

    Format a citation for one paper as BibTeX, RIS or APA 7. The URL is the paper's NobleID ARK when it has a confirmed one, else its DOI. DOIs known to be unreliable are left out rather than guessed.

  • cite

    Citations for several papers at once, in one format, in the order given. Each entry links to the ARK or DOI; failures are listed per identifier instead of failing the whole call.

  • search

    Search scholarly papers and return citable results as {results: [{id, title, url}]}. `url` is open access first. Pass an `id` to fetch for the full text record.

  • fetch

    Fetch one paper as {id, title, text, url, metadata}. `text` is the abstract plus a formatted citation; `url` is open access first.

How to connect

Add the server to your MCP client config. Claude Desktop, Cursor and most other clients read this shape:

{
  "mcpServers": {
    "nobleblocks-mcp": {
      "type": "http",
      "url": "https://mcp.nobleblocks.com/mcp"
    }
  }
}

Add it

Claude Code, in a terminal

claude mcp add --transport http nobleblocks https://mcp.nobleblocks.com/mcp

Add to CursorAdd to VS Code

claude.ai, ChatGPT and other clients: add https://mcp.nobleblocks.com/mcp as a custom connector in the settings.

Each line is something a program found when it went looking: the registry entry, the server's own answer, a spec that parses. A missing line means we did not find it, not that it does not exist.