Codeflowmap shows what AI-written code reads, writes, and protects
Codeflowmap is a tool that scans a and maps how files and functions connect to each other. It focuses on showing where data is read, where data is written, and where -related paths run through the code. It was built for the problem of using to create code that works, but is not fully understood by the maker.
A local model through Ollama, or any , can be connected to add short notes about what each file does and what data it touches. The tool runs locally by default, unless a remote API is connected. Its output can link into an , so the code map can become part of a personal project .
It runs with `bunx codeflowmap serve` and is released under the MIT license.
Key points
- It maps dependencies and call paths across files and functions in a repository.
- It highlights read paths, write paths, and -related paths.
- It can use Ollama or an to explain files and data use.
- It runs locally unless the maker connects it to a hosted API.
- It can link results into an for project notes.