New Project Idea: hidefold

[ata-sesli]

Language-agnostic code minification and obfuscation tool for Octacity

I’m considering building Hidefold as a new Octacity project.

The idea is to create a language-agnostic CLI tool for reducing source-code size and making implementation details harder to understand, while still preserving program behavior.

Hidefold would have two main modes:

• API-preserving minification, where public interfaces remain unchanged while private symbols, formatting, comments, and unnecessary code are minimized
• Full minification and obfuscation, where the tool assumes the entire program is available and can apply more aggressive transformations across module and API boundaries

The tool could eventually support features such as:

• comment and whitespace removal
• private and local symbol renaming
• dead-code elimination
• constant and expression simplification
• string and constant obfuscation
• cross-module symbol rewriting
• configurable preservation rules
• deterministic builds through seeds
• generated symbol maps for debugging
• reports showing size reduction and applied transformations

The long-term goal is not to build a separate obfuscator for every language. Instead, Hidefold would provide a shared transformation engine with language-specific parsers, semantic analyzers, and code generators.

The project would need to handle language-specific concerns such as reflection, serialization fields, exported symbols, FFI boundaries, macros, dynamic imports, and framework-discovered names.

Obfuscation does not make code impossible to reverse engineer, but it can significantly increase the time and effort required to understand, copy, or modify it.

I’d like to hear thoughts about the architecture, useful transformations, which languages should be supported first, and where the boundary between language-independent and language-specific logic should be.

All contributions are welcome.