lacuna-research

n8n community node for Lacuna Research: search and read a knowledge graph of machine learning and AI research (papers, research directions, authors, venues, institutions, hypotheses, and linked code, datasets, and models)

Package Information

Released: this month
Downloads: 0 weekly / 0 monthly
Latest Version: 0.2.0
Author: Tiptree Systems

Documentation

n8n-nodes-lacuna-research

This is an n8n community node. It lets you use Lacuna Research in your n8n workflows.

Lacuna is a knowledge graph of machine learning and AI research. It organizes more than two million papers into tens of thousands of named research directions, links them to authors, venues, and institutions, generates research hypotheses, and tracks the code repositories, datasets, models, and demos that papers publish. This node calls Lacuna's public read-only API, so it needs no account or API key.

n8n is a fair-code licensed workflow automation platform.

Installation
Operations
Credentials
Compatibility
Usage
Resources
Version history

Installation

Follow the installation guide in the n8n community nodes documentation. The npm package name is n8n-nodes-lacuna-research.

Operations

Every operation is a GET request against https://lacuna.tiptreesystems.com/api/v1. Operations that return lists emit one n8n item per element; the others emit one item.

Resource Operation What it returns
Search Search Ranked matches for a free-text query across all entity types, or one type (paper, direction, author, venue, institution, hypothesis, resource). Filters for dates, venue, sort order, ranking profile, lexical fields, resource kind, and provider.
Paper Get The full paper record with abstract and metadata
Paper Get Blog The generated blog-style explainer
Paper Get Concepts Extracted problems, mechanisms, findings, and limitations
Paper Get Context An agent-oriented summary with authors, figures, versions, and links (compact or complete view)
Paper Get Figures Figures with captions and image URLs
Paper Get Neighbors Papers sharing research directions and concepts
Paper Get Preview A short preview card
Paper Get Resources Linked code, datasets, models, and demos
Author Get The full author record
Author Get Context An agent-oriented profile with recent papers and research directions
Author Get Directions Research directions the author has published in
Author Get Impact Per-direction impact scores
Author Get Neighbors Authors whose work is closest
Author Get Papers The author's papers, newest first
Author Get Resources Resources linked to the author's papers
Author Resolve The Lacuna author matching an ORCID, Google Scholar ID, OpenReview profile, DBLP PID, OpenAlex ID, Semantic Scholar ID, homepage, GitHub, or LinkedIn URL
Direction Get The full research direction record with its article and samples
Direction Get Context An agent-oriented summary with top papers, authors, and related directions
Direction Get Papers The papers in the direction
Direction Get Resources Resources linked to papers in the direction
Venue Get Context A venue summary with top directions, authors, representative papers, and activity by year, optionally for one edition year
Institution Get Authors Authors affiliated with the institution
Institution Get Context An institution summary with its most active authors
Hypothesis Get The full generated hypothesis with every version
Hypothesis Get Context A summary with motivation, approach, experiments, and source directions
Resource Get One code repository, dataset, model, or demo with its linked papers
Resource Get Many Resources filtered by query, kind, provider, paper, author, direction, stars, downloads, and update date
Corpus Get Statistics How many papers, directions, authors, venues, institutions, and hypotheses are indexed

Credentials

None. The Lacuna API is public and read-only. Please keep automated traffic reasonable: search calls run a semantic ranking step and take a few seconds each, and the service answers Service temporarily busy when it is overloaded. Use the Retry On Fail setting of the node for scheduled workflows.

Compatibility

Built with @n8n/node-cli and tested against the current n8n release. The node has no runtime dependencies.

Usage

Identifiers or URLs. Every ID field also accepts the URL of the matching Lacuna page. For example, the Paper ID field takes art_a0bd7b7f576e42a18159cc69b1a42233 or https://lacuna.tiptreesystems.com/work/mixture-of-experts-with-expert-choice-routing/version/art_a0bd7b7f576e42a18159cc69b1a42233. Author fields also take the page slug such as hugo-larochelle. Venue fields take the slug such as icml.

Finding IDs. Run a Search first. Each result carries its id and Lacuna url, which you can pass to the Get operations of the matching resource.

Views. Context operations and Direction Get Papers return the compact view, which is sized for AI agents and summaries. Add View under Options and pick Complete for raw records and full lists.

AI agents. The node is marked as usable as a tool, so it can be attached to an AI Agent node. A typical setup is one tool instance for Search → Search with the Query field set to be filled by the model, and one for Paper → Get Context.

Paging. List operations have a Return All switch that fetches every page for you, and a Limit when it is off. Search and Get Neighbors return ranked top results, so they take a Limit only. Search returns at most 50 results per call, with an Offset under Additional Fields.

Resources

Version history

  • 0.1.0: initial release.

Discussion