Quick Setup: aglc add and aglc generate

aglc add — One-shot bootstrap

The add command is the recommended way to set up aglang in any project. It does everything in one step:

aglc add [projectRoot] [--name <ProjectName>] [--out <architecture.ag>]

What it does:

1. Scans your project for manifests (package.json, .csproj, go.mod, Cargo.toml, build.gradle, etc.)
2. Generates a starter architecture.ag spec with discovered components and routes
3. Compiles the spec → architecture.o
4. Installs a git pre-commit hook
5. Emits skill.json for AI agent toolchains

Output:

[aglc add] Scanning /my-project for project manifests...
  ✓ Generated spec → /my-project/architecture.ag
    Components: 3 | Infra nodes: 2 | Contracts: 1
[aglc add] Compiling spec...
  ✓ Compiled → /my-project/architecture.o
[aglc add] Installing git pre-commit hook...
  ✓ Installed hook
  ✓ Emitted skill manifest → /my-project/skill.json

╔══════════════════════════════════════════════════╗
║  aglang setup complete                           ║
╠══════════════════════════════════════════════════╣
║  Next steps:                                     ║
║  1. Open architecture.ag and add invariants      ║
║  2. Re-compile: aglc compile architecture.ag     ║
║  3. Every git commit is now checked              ║
╚══════════════════════════════════════════════════╝

aglc generate — Spec generation only

If you want just the spec without installing the hook:

aglc generate [projectRoot] [--out architecture.ag] [--name MyApp]

This is useful for reviewing the auto-generated spec before committing to it.

After setup: Adding invariants

The generated spec has components and flows — you add invariants to enforce boundaries:

node app_runtime : server { trust: trusted }
node database : postgres { trust: trusted }

component PublicAPI {
  runs_on: app_runtime
  paths: "src/api/**"
}

invariant NoDirectDBAccess {
  deny flow PublicAPI -> database
}

See Language Reference for the full invariant syntax.

For AI Agents

After running aglc add, emit the context file so agents understand the boundaries:

aglc emit-context --arch architecture.o --out AGENTS.md

Agents that support the skill protocol can load skill.json directly. The AGENTS.md file is a plain-English description of all architectural rules suitable for any agent.