> Markdown twin of https://adna.network/learn/concepts/
> Index: https://adna.network/llms.txt · Full corpus: https://adna.network/llms-full.txt
> State is a build-time snapshot generated 2026-08-30 (UTC); nothing here is live.
> Derived from the rendered page — this route has no markdown source.

# Concepts

13 core concepts that make up the aDNA knowledge architecture, ordered from concrete foundations to philosophical principles.

[The Triad The triad is aDNA's universal organizing principle: every piece of project knowledge belongs in exactly one of three directories — what/, how/, or…](/learn/concepts/triad)[The Ontology The aDNA ontology is a typed vocabulary of 16 base entity types — organized across the triad — that defines what kinds of things a project can…](/learn/concepts/ontology)[The Knowledge Graph An aDNA vault is not a filing cabinet — it's a knowledge graph. Files are nodes, wikilinks are edges, and AGENTS.md files are the navigation layer…](/learn/concepts/knowledge-graph)[Governance Files Every aDNA project has five ALLCAPS governance files at its root — CLAUDE.md, MANIFEST.md, STATE.md, AGENTS.md, and README.md. Together, they form the…](/learn/concepts/governance-files)[Token Selection Token selection is the discipline of choosing which knowledge to load into an AI agent's context window — and, critically, which knowledge to leave…](/learn/concepts/token-selection)[The Convergence Model A project can know more than any AI agent can hold in mind at once. The convergence model solves that by narrowing the knowledge in play at each stage…](/learn/concepts/convergence)[Dual Audience aDNA's communication discipline: every content file stays technically precise for developers and genuinely clear for newcomers — neither audience sacrificed.](/learn/concepts/dual-audience)[Context Optimization Context optimization is the practice of designing context files — the curated knowledge agents load before doing work — so they deliver maximum…](/learn/concepts/context-optimization)[Lattice Composition Big jobs are usually too big for one workflow. Lattice composition is how aDNA snaps smaller workflows together to make bigger ones — the same way a…](/learn/concepts/lattice-composition)[Open Standard aDNA is an open standard — a publicly documented specification that anyone can implement, extend, and build upon without permission or payment. The…](/learn/concepts/open-standard)[Agentic Literacy Agentic literacy is the ability to work effectively with AI agents — not just prompting them, but structuring knowledge so agents can find…](/learn/concepts/agentic-literacy)[Context Commons Think of the Context Commons as a shared library of "how to teach an AI assistant your project" — like GitHub, but for agent knowledge instead of…](/learn/concepts/context-commons)[FAIR Metadata FAIR is a simple four-question test: can someone else Find your work, Access it, Interoperate with it, and Reuse it? aDNA bakes that test into every…](/learn/concepts/fair-metadata)