Nobody writes a new browser for every website.
So why does every AI agent need a custom integration for every application?
From Still 2.0, a short film by AtomGraph.
Applications ship their own stylesheets as packages. They are compiled together with the platform's, so an application's templates run in the browser as well.
https://t.co/BIaY077dEs
In many web apps, the interface is built twice: server templates for the first page, front-end components for the browser.
LinkedDataHub 6 renders both with the same XSLT 3.0 stylesheets, Saxon on the server and SaxonJS in the browser. There is no separate front end.
📦 Software Spotlight: LinkedDataHub
Describe what you want to do with linked data in everyday language. Its AI assistant turns your request into a plan you can review and choose to run.
https://t.co/LLcQ5rwRul
Nobody writes a new browser for every website.
So why does every AI agent need a custom integration for every application?
From Still 2.0, a short film by AtomGraph.
Wikipedia is the largest encyclopedia ever written by humans—and it is likely to hold that title for ever. (Any future rival is sure to be made with the help of AI) https://t.co/SuzVVR6gQU
Illustration: Carl Godfrey
The assistant doesn't generate code. It produces a plan of data operations, executed with the user's access rights.
The resulting chart becomes a block on the page, and the conversation is saved in the document too.
https://t.co/VEAOLSELbg
In LinkedDataHub 6, the assistant is available on any page.
Ask in plain language. Read-only requests run immediately. Here, it finds the five customers who ordered the most and presents them as text and a chart.
Changes to data wait until Execute is pressed.
"The data from the ontology itself, is stored as a knowledge graph, persisted to QuipuDB, our homegrown RDF graph database [...] Data here is generated by parsing raw telemetry, logs, and incident discussions into triples (subject-predicate-object facts)."
https://t.co/Lqsprt2y8D
Nobody writes a browser per website. Web apps could work that way.
First Principles of the Web derives a generic application architecture and finds its parts already standardised: RDF, SPARQL, XSLT, CSS.
LinkedDataHub is the reference implementation.
https://t.co/QBVExWyGHf
New on our Frontier Red Team blog: using Claude to formally verify software at large scale.
Formal verification can prove that code is free of certain bugs, but historically it has only been feasible for small codebases (or required a lot of expert time).
With AI models, however, this is beginning to change. In collaboration with @theoremlabs, we used Claude to prove security properties in over one hundred real-world open-source libraries over the span of a few weeks.
Read our research in full here: https://t.co/zUyx8CjBEk
LinkedDataHub 6 is out.
The application is data: classes, forms, views, charts and access rules are RDF documents. You deploy an app by pushing a directory of RDF documents and the files they use, and an agent can write one.
What changed:
https://t.co/BIaY077dEs
Two public sites already run 6.2.2: https://t.co/jngvh6szkW and https://t.co/jGznYe4YV8.
To run it yourself: clone the tag, make up.
https://t.co/pu5KkOpgEv
LinkedDataHub 6 is out.
The application is data: classes, forms, views, charts and access rules are RDF documents. You deploy an app by pushing a directory of RDF documents and the files they use, and an agent can write one.
What changed:
https://t.co/BIaY077dEs
Agents use the same API humans do.
Every dataspace serves /AGENTS.md. The ldh CLI and its Claude Code plugin give an agent the same operations a person has, on the same documents, through the same interface.
History is git.
Turn on versioning and every write is a commit, authored with the WebID of the agent that made it, human or AI. Every versioned document has a TimeMap, a TimeGate and past versions over Memento (RFC 7089).