Your databases, APIs, data lakes and spreadsheets each describe part of your domain, and none of them agree. In a Folio workspace, you and your agents model your domain once and bind that model to each source where the data already lives. Start with one process or one system, and grow the model toward the whole enterprise.
Your model, a query and your agent, in one window
Your model describes your domain's entities, how they relate, and the state, actions and events of the systems behind them. Bound to your sources, the model becomes a digital twin of your enterprise: one graph across your sources. Folio validates the graph against your model and serves it to your agents to query.
- Your model
crm.ttldescribes accounts and deals and binds them to a Postgres CRM.gleif.ttlbinds parent companies from the API of the Global Legal Entity Identifier Foundation (GLEIF). GLEIF records a company's ultimate parent when the company reports one.- A query
pipeline-by-parent.rqjoins the two sources and rolls open deals up to each ultimate parent.- Your agent
- Claude, in Folio's terminal, asks the same graph over the Model Context Protocol (MCP), the protocol AI agents use to call tools.
Quickstart
Welcome to Folio Install Folio, open a folder as a workspace, and meet the views you use to model your domain and query it across sources. Start →Connect your agent Switch on Folio's MCP server, then connect any agent that speaks MCP over HTTP to your workspace, from Folio's own terminal or from outside Folio. Start →Model your first class Define a PowerPlant class and a SolarInstallation subclass of it, write an installation with no name, and watch Folio flag the name it inherits. Start →Publish and install a package Package a domain model as a semantic package, publish a version of it to an S3-compatible bucket, and install a published version in another workspace by its IRI. Start →Query your data with SPARQL Write SPARQL queries over New York City's taxi zones and their green taxi pickups. List the zones, filter and sort them, join two data files, keep the zones with no pickups, and count and add up by borough. Start →Graph your Markdown Create two Markdown files in Folio, pick a mapping, and each becomes an entity in your workspace's graph. Add fields that link them, then query across the graph. Start →
Authoring
Write Turtle Write a class schema in Folio's Turtle Editor: pick a profile, let typeahead fill each slot, fix the warnings the profile raises, and read the file back. Start →Form view Edit a class schema as a form: change a field, add one, and read the ordinary Turtle each change writes back. Start →Graph View How Graph View draws a file: one node per resource, with its types, values, lists, blocks and property shapes drawn inside it. Start →Markdown syntax What each Markdown element means to a mapping, Folio's special syntax, and the mappings for Markdown documents. Start →
Modeling
Classes Each construct a class schema supports, with a Turtle example of what it says about your domain: class, property, required and cardinality, datatype, class of a value, allowed values, value range, pattern, default value, label and comment, and classical and compositional inheritance. Start →Open world and scope Folio keeps RDF's single global scope and uses datasets and profiles to decide what to put in front of you. Start →How profiles work What a profile is, where a file gets one, what the Profile menu shows and changes, what a profile checks, how Markdown files read theirs, and how to write your own. Start →Built-in profiles Each built-in profile, what it checks, and a Turtle or Markdown example: class schemas, RDFS and SKOS vocabularies, datasets, graph views, MCP tools, RML and Markdown mappings, Web of Things descriptions and authoring profiles. Start →Inference The conclusions Folio draws by default: the RDFS and OWL rules for subclasses, subproperties, datatypes, domain and range, equivalent classes and properties and inverse properties, and Folio's own inherited properties, dataset membership and default values, plus what Folio does not infer. Start →Taxonomies Write a hierarchy of concepts in SKOS and Turtle: definitions, alternative and hidden names, labels in other languages, and the codes other systems already use. Pick the SKOS Taxonomy profile and Folio checks it as you type. Start →
Federation
Federation intro Your domain model lives in your workspace and your data stays in the systems that hold it. Folio reads each source through a mapping when a query asks, and nothing is copied. Start →Query a JSON API Describe a web API, map its JSON onto entities in your domain model, and query it live. Start →Query CSV, Parquet and Excel files Model the rows of a CSV file, bind the domain model to the file, and query it. Then read the same file from its URL, and join it with a Parquet file. Start →Query a database Connect Folio to a CRM in Postgres, map its accounts and deals, and join the open pipeline to each account's parent company from GLEIF. Start →Add credentials Sign a data source in without writing a secret into a file. Bind each secret to the OS keychain, session memory, an environment variable, 1Password or an OAuth sign-in, and Folio sends it only to the address you confirmed. Start →