sqlib documentation

Start with concepts.md. It defines Library, Query, QueryGroup, DataGraph, TupleSet, ArgumentSet, RuleSet and the rest of the vocabulary that every other page assumes.

Guides — doing something

• guides/running-and-configuring.md — how do I start a server, and which knobs decide what it does?
• guides/rest-api-walkthrough.md — how do I create a library, save a parameterised query and run it over HTTP?
• guides/query-groups.md — how do I chain queries on the canvas and feed one's results into the next?
• guides/notebooks.md — how do I write a runnable story about a library, with prose between the queries and each result named?
• guides/rules-and-srl.md — how do I write SPARQL RL inference rules and run them against a graph?
• guides/mcp-clients.md — how do I point an MCP client at the server and what can it then do?
• guides/publishing-tools.md — how do I publish a library's queries as an application's MCP tools and HTTP routes?
• guides/mcp-app.md — how do I edit, run and save a query from inside a chat client, how do I teach SPARQL rules there with a library laid out as lessons, and how do I test that locally?
• guides/static-export.md — how do I export a library as a self-contained page that runs without the server?
• guides/testing-and-conformance.md — how do I attach tests to a query or rule set and run a conformance suite?
• guides/etl.md — how do I build RDF from tabular sources with DuckDB SQL?
• guides/prefixes.md — how do I manage namespace prefixes across a library?
• guides/deploying.md — how do I run the container image and the static web app for real?

Reference — looking something up

• reference/configuration.md — every environment variable, its default and its effect.
• reference/feature-flags.md — the feature flags, their env vars, defaults and what each turns off.
• reference/entity-model.md — every entity type, its fields and its relationships.
• reference/srl-language.md — SRL syntax, its constructs and its constraints.
• reference/scripts.md — the package scripts and the scripts under scripts/, and what each one runs.

Explanation — understanding why

• explanation/architecture.md — what the packages are, what runs in which process and how a request travels.
• explanation/versioning-and-immutability.md — why entities split into a stable pointer and immutable versions, and what that buys.
• explanation/environments.md — named version pointers (dev, prod) inside a library, how a run resolves them, and who may move them.
• explanation/storage-and-caching.md — where the library actually lives, and what the in-memory cache is for.
• explanation/security-model.md — what the auth layer enforces, what it does not, and why some doors are open.
• explanation/rdf-patch.md — how an update query's effect is derived before it is applied.
• explanation/inline-inputs-and-the-rail.md — where a saved input lives, why pasting a graph into a call saves it as a graph, and what the rail shows about where things came from.

Examples

• examples/sudoku-solver-ruleset.md — a worked rule set that solves a Sudoku by inference, end to end.

Contributor setup, the CI checks and commit conventions are in CONTRIBUTING.md. Reporting a vulnerability and the deployment security posture are in SECURITY.md.