ng-galien/code-moniker

code-moniker is a tool that analyzes source code and builds a symbol graph, a structured map of every meaningful identifier in a codebase and how those identifiers relate to each other. Rather than treating code as plain text, where a search might match the wrong thing, it parses each file and produces a stable identity for every function, class, variable, and module. These identities are called monikers, URI-like paths that encode not just the name of a symbol but what kind of thing it is and where it lives in the project's folder structure. The tool works in two modes. A command-line interface lets you inspect symbols, filter them by type, and run a linter that enforces architecture rules, for example requiring that code in a domain folder never imports code from an infrastructure folder. These rules can run before a commit, after every file edit, or inside a continuous integration pipeline. The second mode is a PostgreSQL extension: it installs custom data types into a database so you can store and query the same symbol graph using SQL, which makes it easy to join symbol data with other project information. Supported languages include TypeScript, JavaScript, Rust, Java, Python, Go, C#, and SQL. The project itself is written in Rust and uses pgrx to build the database extension, along with tree-sitter for parsing source files. You would reach for code-moniker when text search is too imprecise, such as finding every class that implements a particular interface, checking whether a layered architecture is being respected, or building an AI coding agent that needs accurate, structured knowledge of a codebase rather than raw text. Rule expressions are written in a small DSL and configured in a .code-moniker.toml file that merges on top of built in defaults. The project is dual licensed under MIT or Apache 2.0, at the user's option.