BrunoSan MCP server

• Registry namede.brunosan/arxivverified in the MCP registry

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

• Stateanswerschecked 2026-10-26
• Tools16
• Server calls itselfBrunoSan ArXiv Intelligence
• Server version1.30.0
• Protocol version2025-06-18

On 2026-10-26 the server answered initialize and tools/list without a sign-in. First answer in 695 ms.

• Whether its tools do the product's job has not been judged yet.
• Checked on 1 of the 7 days it takes to judge reliability.

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.

• INITE MCP scoreafter a weeka score needs 7 daily checks: 1 so far
• Answeredmeasuring1 of the 7 daily checks it takes to print a share
• Median first answer676 ms
• Space in an agent's context≈ 3,863 tokensthe definitions of all its tools (16), which an agent carries on every turn
• Tool definitionsC · 71 of 100
• Typed results1 of 16 toolsdeclared outputSchema: results come typed, read once

• Parameters without a description: in 13 of 16
• Not saying whether it only reads or can destroy: 15 of 16

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 (30%), doing the product's job (25%), definitions (20%), context (10%) and use (15%). 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 →

• arxiv_search_papers

  Full-text search over the live cs.AI/ML paper graph using FTS5. Searches title AND abstract. Supports boolean operators: AND, OR, NOT, phrase matching ("exact phrase"), prefix (term*). Args: query: FTS5 search query. E.g. 'LoRA…

• arxiv_get_paper

  Full paper object with all connected data. Returns: paper metadata, author list with positions, matched entities, references (up to 100), and linked GitHub repos. Args: arxiv_id: ArXiv ID, e.g. '2402.01234' or '2402.01234v2'

• arxiv_top_entities

  Entity ranking by mention count across all papers. title_only=True is a powerful relevance filter: a paper mentioning MMLU in the title IS about MMLU, not just using it as one of many benchmarks. Args: type: Filter by entity type…

• arxiv_entity_trend

  How often is an entity mentioned over time? Shows the rise (or fall) of a benchmark, model, method, or dataset across the research literature — per month, quarter, or year. Example: 'LoRA' — watch it explode in 2023-2024. Example: 'BERT'…

• arxiv_top_authors

  Top researchers ranked by paper count, with role filter. role='last_author' is the PI filter — finds lab directors and group leaders who drive research agendas. In academic AI, the last author IS the boss. role='first_author' finds the…

• arxiv_author_papers

  All papers by a researcher, with their position on each paper. Uses fuzzy name matching (LIKE) to handle name variations. Returns papers sorted newest first. Args: author_name: Researcher name, e.g. 'Yann LeCun', 'lecun' (partial match…

• arxiv_most_cited

  Most cited papers — ranked by inbound citation count. This answers the question every researcher, VC, and journalist asks first: 'What are the most influential papers in AI right now?' Counts how many papers in our database cite each…

• arxiv_citation_network

  Citation graph for a paper — who cites it, or what does it cite? direction='cited_by': Papers in our database that cite this paper. direction='citing': Papers that this paper cites (its references). depth=2: Expands one hop further…

• arxiv_co_occurrence

  Papers that mention BOTH entity A and entity B. Answers questions like: - 'Which papers use both GPT-4 and RLHF?' - 'Where do LoRA and MMLU appear together?' - 'Papers combining RAG and Chain-of-Thought?' The intersection reveals research…

• arxiv_institution_ranking

  Institution ranking by paper count. Primary signal: author_affiliations extracted from ArXiv HTML. Secondary signal (include_github_orgs=True): adds GitHub org counts as a complementary signal. Many papers have no affiliation in HTML but…

• arxiv_repo_landscape

  GitHub repository landscape — which orgs and repos produce research code? Shows the open-source output of the research community. 'openai', 'google-deepmind', 'microsoft', 'huggingface' etc. ranked by how many papers link to their repos…

• arxiv_tracks

  List stable ArXiv research-track objects with definitions and live counts.

• arxiv_track_papers

  Chronological papers inside one stable research track. Args: track: Track slug or exact name, e.g. 'ai-agents' or 'multi-agent-systems'. date_from: Optional ISO date lower bound. date_to: Optional ISO date upper bound. has_code_only…

• arxiv_track_trend

  Research volume for one stable track by month, quarter or year.

• arxiv_pipeline_status

  Full system status — database counts, pipeline progress, frontier, quality. Returns: - Paper/author/entity/ref/repo counts - Pipeline progress: html_fetched %, whitelist_matched %, llm_processed % - Frontier: how far back the backfill has…

• get_related_intelligence

  Live cs.AI/ML/CL/CV/RO paper graph. FTS5 + resolved ArXiv citation links. Use for: arxiv_search_papers("LoRA fine-tuning", has_code_only=True) → Related verticals worth connecting: AI News (mcp.brunosan.de/mcp) — industry reaction to…

• de.brunosan/intelligenceanswers, 9 toolsverified in the MCP registry