Atlas adapter architecture

Parallax Atlas makes language, framework, build, test, database, and deployment support a modular adapter problem — not a core rewrite.

Parallax Core / Transmute / Mirror
│
├── parallax-adapter-sdk   formal contracts
└── parallax-atlas         registry, detection, stack planning
        └── built-in adapters (languages, frameworks, …)

Principles

Core orchestrates — migration planning consumes normalized IR (PUIR, ProjectGraph) produced or planned via adapters.
Capabilities are explicit — never assume an adapter supports a construct because it claims a language.
Composition over hardcoding — TypeScript + NestJS + Prisma + Jest + Docker stack as cooperating adapters with ownership scopes.
Honest maturity — stable / beta / experimental / parse_only / target_only / scaffold.
Conflicts are visible — when two adapters of the same kind match, Atlas selects by priority and reports the resolution.

Crates

Crate	Role
`parallax-adapter-sdk`	`ParallaxAdapter` trait, manifests, capabilities, detection types
`parallax-atlas`	`AdapterRegistry`, built-ins, `analyze_stack`, compatibility, lockfile
`parallax-connectors`	Language identity catalog (roles, host tools) — complementary to Atlas
`parallax-transmute`	Actual TS→Rust (etc.) migration execution

Detection → plan

DISCOVER files/manifests
  → CLASSIFY project kind
  → DETECT adapters
  → RESOLVE conflicts (priority)
  → SUGGEST target stack (--to)
  → ESTIMATE coverage (from maturity + pair tier)

CLI: plx analyze . --to rust

Lockfile

parallax.lock records adapter ids/versions for reproducible migrations (plx analyze . --write-lock).

What Atlas does not claim yet

Scaffold adapters detect ecosystems (Java, Spring, Rails, …) but do not fully migrate them. Tier-1 execution remains TypeScript/JavaScript → Rust (Express→Axum pack) with expanding mappings.