Architecture

Architecture

KaspaScript is organized as a compiler pipeline with a programmability kernel on top.

Pipeline

Source.ks
  -> lexer
  -> parser
  -> AST
  -> semantic analysis
  -> typed IR
  -> Kaspa txscript backend
  -> deterministic artifact JSON
  -> kernel package JSON

Workspace Crates

Crate	Purpose
compiler/lexer	Position-tagged token stream.
compiler/parser	Contract AST for params, spend paths, expressions, arrays, calls, and fields.
compiler/semantic	Type, scope, finality, builtin, and target-safety checks.
compiler/ir	Opcode-agnostic typed instruction layer.
compiler/codegen	Source-grounded Kaspa txscript bytecode emission and verification.
compiler/protocol	Protocol target manifests and feature gates.
kernel	Contract blueprints, capability profiles, wallet previews, indexer schema, fee policy, and readiness.
sdk	Compile API and testnet harness surfaces.
cli	Developer commands for compile, inspect, verify, package, wallet, tx, and proof flows.

Design Principles

• Deterministic artifacts.
• No hidden fees or hidden spend behavior.
• Unsupported behavior fails before bytecode emission.
• Source-grounded protocol claims.
• Separate targets for verified, preview, TN10 Toccata, and future mainnet.
• Mainnet remains blocked until activation evidence is pinned.