Adaption
Import, adapt and expand training datasets, then launch AutoScientist fine-tuning runs from an AI agent.
Search ML and AI papers, authors, venues and research directions from Claude.
Opens Agent Studio, where connecting is one click. The connector URL below works in any MCP client.
Paste it into any MCP client. Setup docs
The Lacuna Research MCP server, from Tiptree Systems, lets Claude search a knowledge graph of machine learning and AI research: papers, authors, venues, institutions and research directions. Its 13 tools are all annotated read-only, it needs no account or sign-in, and it suits literature reviews, related-work sections and prior-work checks.
Verified connector
Listed by Anthropic as a partner connector in its Connectors Directory.
Connection checked by Agentman on .
Anthropic states this reflects the level of review a connector received, not a security audit.
Search Lacuna's ML/AI corpus for papers, research directions, authors, venues, institutions, novel research hypotheses, and research resources (code repositories, datasets, models, and demos linked to papers). Start here for any question without an ID. Set search_type to what the question is about: paper for literature; cluster for a research direction ("who works on X", "what is the state of X"); author, institution or venue for a person, a lab or a conference; hypothesis for open research ideas; resource for code, datasets and models. Then pass a result's id or url to the matching detail tool. A query that is an arXiv ID, a DOI, or an arXiv, DOI or OpenReview URL returns that exact paper. Paper results carry `identifiers` (arXiv ID, DOI, OpenReview URL where known), which other research tools accept. For novel ML/AI research ideas, use search_type="hypothesis", then get_hypothesis on promising results. search_type accepts all, cluster, paper, author, institution, venue, hypothesis, or resource. Use paper for literature and get_paper to read a result. Use other sources for biographies, news, and non-research web content. Resources: search_type="resource" searches GitHub repositories, Hugging Face datasets/models, and Zenodo records that Lacuna has linked to at least one paper. To find benchmark or evaluation datasets, use search_type="resource" with resource_kind="dataset" (for example query "image classification benchmark" or a dataset name like "ImageNet"). resource_kind accepts codebase, dataset, model, or demo (one value or a list); provider accepts github, huggingface, or zenodo. Both filters require search_type "resource" or "all" (with either filter set, only resources are returned). Each result has an external `url`, a Lacuna `context_url`, and `linked_paper_count`; call get_resource on its id to list the linked papers (with their relationship to the resource), then get_paper on those papers for reported results and numbers. Resource search does not support date_from/date_to, venue, year sorting, or semantic ranking. ranking_profile accepts: - default / lexical (default): production ranking; relevance-sorted paper searches combine lexical and semantic retrieval when fields is unset. - semantic: conceptual paper retrieval; supported for paper and all. - bm25_title_abstract / bm25: lexical paper matching over those fields. sort accepts relevance (default), year_desc, or year_asc. Semantic ranking cannot use year sorting; constrain recency with date_from/date_to instead. date_from and date_to are inclusive YYYY, YYYY-MM, or YYYY-MM-DD bounds. author_id_or_url constrains a paper search to one author. Pass an author ID or Lacuna author page URL; it requires search_type="paper". fields optionally restricts and weights lexical fields, for example "title^4,abstract". Supported names are title, abstract, summary, concepts, name, top_names, and venue, plus description, topics, provider_key, and paper_titles for resources (a resource's title already includes its description, topics, README text, and linked paper titles). Fields must exist on the selected search_type, weights must be within 0 < weight <= 100, and fields cannot be combined with semantic ranking.
Fetch a generated novel ML/AI research proposal from Lacuna. Use after search_lacuna(search_type="hypothesis") to inspect a proposal. view selects the response shape: - "context" (default, recommended): compact single-fetch proposal context — summary_markdown, abstract, and linked directions, with the raw upstream record (whose markdown duplicates summary_markdown) dropped server-side. - "full": the server's version record, including version history and signal counts. Proposal bodies are in versions[].markdown. Use only when you need versions or signals.
Describe one research direction: what it is, who leads it, what is nearby. Use for "who works on X", "what is the state of X" and "what is adjacent to X", after search_lacuna(search_type="cluster") finds the direction. The context names the direction's leading authors, key papers and related directions. Use a numeric cluster ID (for example 25005) or a direction URL ending in that number. view selects the response shape (`context` typically contains the fields `full` provides plus the agent-oriented summary content): - "context" (default, recommended): compact agent-oriented summary with summary_markdown, capped papers/authors/related_directions, and truncation markers. - "full": raw upstream cluster record only. Cheaper than context when you only need basic cluster metadata.
Fetch paginated papers associated with a Lacuna research direction/cluster. view selects the per-paper shape: - "compact" (default, recommended): citation-ready rows (id, url, title, year, venue, a few authors, abstract snippet). Drops the raw upstream info blob and levels.cluster internals that otherwise dominate the payload. - "full": the complete upstream paper records. Much larger; use only when you need the raw metadata (openalex/dblp/arxiv ids, etc.).
Fetch one paper: its summary, authors, venue, figures, linked resources, or full text. Accepts a Lacuna artifact ID or paper URL from search_lacuna, or an outside identifier for the same paper: an arXiv ID (2310.06825), a DOI, or an arXiv, DOI or OpenReview URL. Use this when the user names a specific paper or after a search. For an outside identifier that is not in Lacuna's ML/AI corpus, the result is `found: false` with a note saying so. The default view is a summary; ask for the full text only when the summary does not answer the question. include_resources adds linked resources (code repositories, datasets, models, and demos) in context and full views by default. Pass False to omit them; other views ignore this option. Pass a resource entry's id to get_resource for further details. view selects the response shape. `context` requests Lacuna's compact agent-oriented context by default, while the four single-field views (blog/figures/concepts/neighbors) return isolated sub-resources: - "context" (default, recommended): agent-oriented summary with summary_markdown, authors, and a small figure preview. Start here for almost everything. Available versions are included when present; to read another version, use its artifact ID from the returned version list. - "full": raw upstream paper record. Cheaper than context when you only need basic metadata. - "preview": compact card with unique fields `excerpt`, `excerpt_kind`, `bookmarked`. Use for citation-style display. - "blog": just the summary_markdown content, without the rest of the context envelope. - "figures": just the figures list. - "concepts": just the concepts list. - "neighbors": just the related-papers list. - "fulltext": the paper itself, `max_chars` at a time (default 40,000, at most 100,000) from `offset`; pass the returned `next_offset` to keep reading. `format` says what you got: "latex" is the arXiv LaTeX source without its bibliography (citations appear as \cite keys); "pdf_text" is text extracted from the PDF, where equations and tables can be garbled. A scanned paper has no text. figure_limit (context view only) caps the figure preview (server default 3). Pass 0 to suppress figure previews while keeping a `figures_truncated` signal.
Fetch a Lacuna research resource (code repository, dataset, model, or demo). Accepts a resource artifact ID (art_...) or a Lacuna resource URL containing that ID (/resource/<slug>/art_...). The response includes the external url, a summary, facets (tasks, modalities, size, license, access), provider metrics, a README/card excerpt, `publications` (linked papers with paper_id, title, venue, year, and relationship such as dataset_for or code_for), related research directions, and other resources mentioned by this one. Pass a publication's paper_id to get_paper to read the paper and its reported results.
List what one author has published, newest first, one page at a time. Use for "what has X worked on recently". Pass an author id or URL from search_lacuna(search_type="author") or from any paper's author list.
List the research directions one author publishes in, one page at a time. Use for "what areas does X work in" or to check whether a person is active in a given direction. Pass an author id or URL.
Profile one researcher: their main papers and the directions they shaped. Use for "who is X in research terms" or "what is X known for". Author profiles describe research output (papers, directions, impact). A free-text `affiliation` field may be present but can be incomplete or outdated and may not represent current employment; the corpus has no biography or employment history, so do not infer those from this data. view selects the response shape: - "context" (default, recommended): Lacuna's compact author context — capped readable papers and an `impact_directions` list (named research directions) in place of the raw numeric `impact_clusters` telemetry, with the duplicated nested author record dropped server-side. - "full": the bounded full-shape author context (raw `impact_clusters`, nested author record; server collections are capped at 100). Set include_neighbors=True to include similar authors as named, linkable records. Neighbor computation may add significant server latency.
List researchers whose work is closest to one author's, ranked. Use for "who else works on what X works on": finding possible collaborators, reviewers or related groups. Pass an author id or URL.
Describe a conference or journal: its top authors and directions, by year. Use for "what did NeurIPS 2025 focus on" or "who publishes most at X". Venue keys are opaque hashes (e.g. "d7bf22905bd6"), never human-readable names like "icml". Find the key first via search_lacuna(search_type="venue") or pass a /venue/... page URL. view selects the response shape: - "context" (default, recommended): compact venue context — capped top authors, non-placeholder top clusters, and a recent-activity slice (the requested `year` is always included), with the duplicated venue block and full year histogram dropped server-side. - "full": the complete venue context (full year histogram, all top authors/ clusters, duplicated venue record).
Describe the research output of a university, lab or company. Use for "what does X publish" or "who are the most active authors at X". The context summarizes the institution's research output: its paper and author counts and the authors who publish most often under it. Affiliations come from the author affiliations listed on published papers, not from staff directories, and may be incomplete or out of date. The response has no contact details, biographies or employment records. view selects the response shape: - "context" (default, recommended): compact institution context — capped top authors with the duplicated institution block dropped server-side. - "full": the complete institution context shape (duplicated institution record and a bounded author list with explicit truncation metadata). Use get_institution_authors to page through the complete author list.
Fetch one page of authors who list a Lacuna institution on their papers. Each author appears because published papers list this institution as one of their affiliations. This is a record of research output, not a staff directory: it may be incomplete or out of date and does not show current employment. Each entry has the author's name, Lacuna page and paper count; there are no contact details or other personal data. Results are ordered by paper count. Use authors_total, authors_returned, authors_offset, and authors_truncated to page through the complete list.
Read from the server on 2026-11-18, including each tool's own safety annotations.
llms.txt says content is generated from extracted concepts, not raw paper text, and the paper_context prompt tells the assistant to attribute a summary's numbers to Lacuna unless it checks the full text.initialize, tools/list, prompts/list, prompts/get and resources/list. Claims about behaviour come from schemas, annotations, prompt bodies and Tiptree's documentation.initialize, tools/list, prompts/list, prompts/get and resources/list against https://lacuna.tiptreesystems.com/api/mcp (2026-11-18), anonymous, HTTP 200. Source for the 13 tools, every schema and annotation claim, the four prompts and their bodies, the empty resources list, the server instructions and serverInfo. Schemas cached at docs/marketing/data/tool-schemas/lacuna-research.tools.json. · retrieved 2026-11-18documentation value (read 2026-11-18; HTTP 200). Source for setup, the no-account statement, the local install, the terms reference and the "thin adapter" description. · retrieved 2026-11-18llms.txt (read 2026-11-18; HTTP 200). Source for the generated-content note and the public API and Markdown routes. · retrieved 2026-11-18docs/marketing/data/anthropic-directory-2026-11-17.json. Source for the partner tier, the 2026-10-13 listing date, the corpus figures, the tool and prompt names and the client list. · retrieved 2026-11-17Connect Lacuna Research once and your agents call these tools on their own: on a schedule, in a workflow, with nobody at the keyboard.