E11 Alloy

Appearance

E11 Alloy ​

Knowledge as a build artifact.

Alloy is a versioned library of markdown documents that teach AI agents how to build software a particular way, plus the pipeline that produces, validates, measures, and distributes those documents. The thesis is simple: treat knowledge the way a compiler treats code. Every doc is machine-validated before it can merge, stamped with the exact commit it was derived from, mechanically declared stale when that commit's files change, and served to agents through a versioned contract they pin like a dependency.

🚀 New here? Start with Why Alloy - what it is, and why a docs folder is not enough. Two minutes.

This guide is for a team standing Alloy up on their own repo: onboarding a codebase Alloy has never seen, establishing a corpus specific to it, and keeping that corpus honest as the code moves. You consume the tooling as published packages, and you drive it through your AI coding agent - the agent runs the commands; you review and commit.

Start here ​

Adopting Alloy on your repo? This is the path. → Quick start - one bootstrap command, then your agent does the rest: survey, review, campaign, and the console at the end. One sitting.

Want the reasoning behind each step? → The full walkthrough - the same journey with the depth: what each artifact means, how to judge the draft, what the machinery guarantees.

Never installed the tooling? → Prerequisites - registry access (Cloudsmith for clients, GitHub Packages for E11 engineers), Node, and an agent. Then Installation for the consumer paths.

Want the vocabulary first? → Glossary - source, unit, kind, view, provenance, drift, coverage, dependency, curated.

Want the mechanism, not the how-to? → Concepts - the two loops, discovery, extraction and the gate, provenance and drift, views and coverage.

Table of contents ​

Getting started ​

The full walkthrough - stand up Alloy on your repo ​

Concepts ​

Reference ​

E11 Alloy - knowledge as a build artifact.

Part of the Engineering11 platform.