READ.ME

A Semantic Knowledge Graph Engine

GitHub – neurons-me/.me: Here we’re codependently creating .me while it concurrently creates us.

Real Performance

.me uses true O(K) reactivity — when a value changes, only its actual dependents update. Not the whole graph.

The Equations

1 million nodes in memory1 sensor changed → exactly 6 dependent nodes recomputedTime to propagate: 0.256msK=6 out of 1,000,000 — the rest of the graph is untouched

Scale the graph to 10 million nodes — if your change has 6 dependents, it still takes the same time.

Syntax: Data that thinks. Logic that explains itself.

.me uses an infinite proxy — any path you write becomes a node in the graph.

No schema. No migrations. No declarations upfront.

me.city.population = 700_000
me.city.name = “Veracruz”
// derived – recomputes automatically
me.city.density = () => me.city.population / me.city.area
// context-aware
me.robot.canProceed = () => me.robot.canLift && !me.robot.needsHumanReview
// stealth – structurally invisible to outside observers
me.wallet[“_”].balance = 1000
// explain any value
me.explain(“city.density”)
// → { value: 3500, expression: “population / area”, dependsOn: […] }
// query across the graph
me.robots[r => r.canProceed === true].namehttps://medium.com/media/424fe6c0bef9f901c8e1148fad4ea9df/href

The Tiny Mental Model

A shop in .me is just a path:

me.shops[1].name(“Downtown”)
me.shops[1].menu.latte(4.5)
me.shops[1].menu.espresso(3.0)

Step 1: Seed the Shops

me.shops[1].name(“Downtown”)
me.shops[1].menu.latte(4.5)
me.shops[1].menu.espresso(3.0)me.shops[2].name(“Riverside”)
me.shops[2].menu.latte(5.0)
me.shops[2].menu.espresso(3.5)me.shops[3].name(“Station”)
me.shops[3].menu.latte(4.8)
me.shops[3].menu.espresso(3.5)

Three shops, two prices each. Nothing smart yet — just facts.

Step 2: Derive Deals and Policies Across All Shops

Apply rules to every shop’s menu at once with [i]:

me.shops[“[i]”].menu[“=”](“breakfastDeal”, “latte + espresso – 1.5”)
me.shops[“[i]”].menu[“=”](“isPremium”, “breakfastDeal > 6.5”)

Now query across all of them:

me(“shops[1..3].menu.breakfastDeal”)
// → { 1: 6, 2: 7, 3: 6.8 }me(“shops[menu.isPremium == true].name”)
// → { 2: “Riverside”, 3: “Station” }

The filter runs over live derived values. No SQL, no query builder.

Demos

Robots that Understand Context — Same physical object, radically different meaning and behavior per robot.

Robots That Understand Context

https://neurons-me.github.io/.me/docs/Splitting-your-Bill.html

Splitting your Bill

Hemisphere Scale — 1 million nodes with cross-domain reactive updates.

Hemisphere Scale

Extreme Fan-Out — One write instantly updates 100k dependents.

View all demos →

Write anything. Chain anything. The kernel figures out the dependencies.

If it changes, everything that depends on it updates — automatically.

https://medium.com/media/1dbcd2a150d8156ec0e31ee71369d6ac/hrefREAD.MEneurons-me – Overviewthis.me

READ.ME was originally published in Coinmonks on Medium, where people are continuing the conversation by highlighting and responding to this story.