The vertical · built on SIS

Starlight Knowledge Tree

An open intelligence graph for human capability, scientific knowledge, and contribution paths.

It maps what exists, what matters, what you know, what you can build, what you can contribute, and what unlocks next.

Explore the Tree →Contribute on GitHub Read the Ontology

The thesis

Track capability, not consumption

Most learning systems track what you've consumed — videos watched, courses completed, pages read. Starlight Knowledge Tree tracks what you can do. The unit is capability, and capability is earned along one loop:

Concept
Skill
Practice
Artifact
Evidence
Contribution
Identity

Identity loops back into concepts: who you become reshapes what you learn next. The graph is the map of that loop, run at the scale of all human knowledge.

Built on SIS / SIP

Substrate, then vertical

The Knowledge Tree doesn't replace the Starlight Intelligence System — it stands on it. The system is the substrate; the tree is a vertical that grows from it.

Substrate · SIS / SIP

Protocol, memory, vaults, MCP tools, and agent-readable infrastructure.

Protocol — the SIP contract for sovereign, attested composition
Memory — persistent vaults and durable state
MCP tools — agent-readable infrastructure
Agents — the council that reasons over the graph

Read the protocol →

Vertical · Knowledge Tree

Research maps, skill trees, open problems, contribution quests, and progression paths.

Research maps — what exists and what matters
Skill trees — capability, not consumption
Open problems — the frontier, made navigable
Contribution quests — work that compounds
Progression paths — what unlocks next

Four initial paths

Start where the leverage is

Four reference branches, each a route from root concepts to a real contribution. Pick one and the graph shows the sequence.

AI Architect

Design intelligence systems that compose without losing control.

Root concepts

Information theoryDistributed systemsAgent orchestrationContext & memory design

Core skills

System decompositionPrompt & tool designEvaluation & attestationRetrieval architecture

Tools

Claude / MCPVector + graph storesTypeScript / PythonEval harnesses

First build artifact

A working multi-agent system with a memory layer and an eval suite.

Contribution path

Open-source an MCP server or a reusable agent pattern.

Space Builder

Move mass, energy, and information off the surface of a planet.

Root concepts

Orbital mechanicsPropulsion & thermodynamicsMaterials scienceControl theory

Core skills

Trajectory modelingSystems engineeringSimulation & telemetryHardware-in-the-loop testing

Tools

Python / GMATCAD & FEAKSP / poliastroEmbedded toolchains

First build artifact

A simulated mission profile with a verified delta-v budget.

Contribution path

Publish a reproducible simulation or an open hardware design.

Bio / Human Intelligence

Understand and extend the living, thinking substrate.

Root concepts

Molecular biologyGenetics & the genomeNeuroscienceSystems physiology

Core skills

Experimental designBioinformaticsStatistical inferenceLiterature synthesis

Tools

Python / RBiopythonNotebooksPublic datasets (open only)

First build artifact

A reproducible analysis notebook over an open dataset.

Contribution path

Share a dataset, a method, or a validated visual explainer.

Creator-Founder

Turn taste and systems into products people pay for.

Root concepts

Audience & distributionProduct & pricingNarrative & brandUnit economics

Core skills

Content engineeringOffer designFunnel & analyticsShipping cadence

Tools

Next.js / VercelSuno / generative mediaAnalyticsPayments

First build artifact

A shipped product with its first real revenue.

Contribution path

Open-source a template, a playbook, or a teardown.

Root node radar

Some concepts unlock entire branches

Not all knowledge is equal. A few root-node concepts — linear algebra, thermodynamics, the cell, the feedback loop — unlock huge regions of the graph downstream. The Root Node Radar identifies those high-leverage nodes and the shortest learning sequences that reach them.

Instead of a flat list of topics, you get a prioritized path: learn this, and these twelve branches become reachable.

Personal progression

Map your Skill DNA

The public graph maps what exists. Your Skill DNA maps where you stand on it. A forthcoming capability — a personal overlay that situates you on the tree across six dimensions.

Interests

The branches you're pulled toward.

Current abilities

What you can already do, honestly assessed.

Proof artifacts

Evidence — built, shipped, measured.

Desired identity

Who you're becoming, named.

Active quests

The contributions you're working on now.

Next unlocks

High-leverage nodes the graph recommends next.

Skill DNA is personal by design. It lives in your local / private vaults — see the privacy boundary below.

Open contribution layer

Everyone can grow the tree

Researchers, students, builders, and creators extend the graph. Every contribution is a node others can build on.

Papers

Tools

Datasets

Simulations

Experiments

Learning paths

Open-source tasks

Visual explainers

Open a contribution quest →

Privacy boundary

Public knowledge, private self

The line is bright and non-negotiable. Shared knowledge belongs in the public repo. Anything personal stays in local / private vaults — never in the public Knowledge Tree.

Public — belongs in the repo

Concepts, skill trees, and ontology
Open problems and research maps
Tools, datasets, and reproducible methods
Contribution quests and learning paths

Private — local vaults only

Health and biometric signals
Genetic data
Financial records
Identity and personal documents

Health, biometric, genetic, financial, and identity data never leave your machine.

The future of learning is not a course library. It is a living map of human capability.

View GitHub →Open Protocol Start a Path

Built on SIP · A Starlight Intelligence vertical

Track capability, not consumption

Concept
Skill
Practice
Artifact
Evidence
Contribution
Identity

Identity loops back into concepts: who you become reshapes what you learn next. The graph is the map of that loop, run at the scale of all human knowledge.

Substrate, then vertical

The Knowledge Tree doesn't replace the Starlight Intelligence System — it stands on it. The system is the substrate; the tree is a vertical that grows from it.

Substrate · SIS / SIP

Protocol, memory, vaults, MCP tools, and agent-readable infrastructure.

Protocol — the SIP contract for sovereign, attested composition
Memory — persistent vaults and durable state
MCP tools — agent-readable infrastructure
Agents — the council that reasons over the graph

Read the protocol →

Vertical · Knowledge Tree

Research maps, skill trees, open problems, contribution quests, and progression paths.

Research maps — what exists and what matters
Skill trees — capability, not consumption
Open problems — the frontier, made navigable
Contribution quests — work that compounds
Progression paths — what unlocks next

Start where the leverage is

Four reference branches, each a route from root concepts to a real contribution. Pick one and the graph shows the sequence.

AI Architect

Design intelligence systems that compose without losing control.

Root concepts

Information theoryDistributed systemsAgent orchestrationContext & memory design

Core skills

System decompositionPrompt & tool designEvaluation & attestationRetrieval architecture

Tools

Claude / MCPVector + graph storesTypeScript / PythonEval harnesses

First build artifact

A working multi-agent system with a memory layer and an eval suite.

Contribution path

Open-source an MCP server or a reusable agent pattern.

Space Builder

Move mass, energy, and information off the surface of a planet.

Root concepts

Orbital mechanicsPropulsion & thermodynamicsMaterials scienceControl theory

Core skills

Trajectory modelingSystems engineeringSimulation & telemetryHardware-in-the-loop testing

Tools

Python / GMATCAD & FEAKSP / poliastroEmbedded toolchains

First build artifact

A simulated mission profile with a verified delta-v budget.

Contribution path

Publish a reproducible simulation or an open hardware design.

Bio / Human Intelligence

Understand and extend the living, thinking substrate.

Root concepts

Molecular biologyGenetics & the genomeNeuroscienceSystems physiology

Core skills

Experimental designBioinformaticsStatistical inferenceLiterature synthesis

Tools

Python / RBiopythonNotebooksPublic datasets (open only)

First build artifact

A reproducible analysis notebook over an open dataset.

Contribution path

Share a dataset, a method, or a validated visual explainer.

Creator-Founder

Turn taste and systems into products people pay for.

Root concepts

Audience & distributionProduct & pricingNarrative & brandUnit economics

Core skills

Content engineeringOffer designFunnel & analyticsShipping cadence

Tools

Next.js / VercelSuno / generative mediaAnalyticsPayments

First build artifact

A shipped product with its first real revenue.

Contribution path

Open-source a template, a playbook, or a teardown.

Some concepts unlock entire branches

Instead of a flat list of topics, you get a prioritized path: learn this, and these twelve branches become reachable.

Map your Skill DNA

The public graph maps what exists. Your Skill DNA maps where you stand on it. A forthcoming capability — a personal overlay that situates you on the tree across six dimensions.

Interests

The branches you're pulled toward.

Current abilities

What you can already do, honestly assessed.

Proof artifacts

Evidence — built, shipped, measured.

Desired identity

Who you're becoming, named.

Active quests

The contributions you're working on now.

Next unlocks

High-leverage nodes the graph recommends next.

Skill DNA is personal by design. It lives in your local / private vaults — see the privacy boundary below.

Public knowledge, private self

The line is bright and non-negotiable. Shared knowledge belongs in the public repo. Anything personal stays in local / private vaults — never in the public Knowledge Tree.

Public — belongs in the repo

Concepts, skill trees, and ontology
Open problems and research maps
Tools, datasets, and reproducible methods
Contribution quests and learning paths

Private — local vaults only

Health and biometric signals
Genetic data
Financial records
Identity and personal documents

Health, biometric, genetic, financial, and identity data never leave your machine.