I built doksnet, a small Rust CLI that lets you link a documentation section to a code snippet and verify that both sides stay in sync using Blake3 hashes.

You define mappings like:

• README.md:15-25

• src/lib.rs:40-65

doksnet stores the ranges + their hashes in a compact .doks file. doksnet test re-extracts the content and fails (exit code 1) if anything changed — including whitespace.

Basic flow:

• doksnet new – initialize

• doksnet add – create doc ↔ code mapping (interactive)

• doksnet test – CI-safe verification

• doksnet test-interactive – review/fix mismatches

It’s repo-local, no external service, no parsing/AST magic — just deterministic text extraction + hashing.

There’s also a GitHub Action if you want to enforce sync in CI.

Repo: https://github.com/Pulko/doksnet Install: cargo install doksnet Web: https://doksnet.pulko-app.com

Interested in feedback on the approach — especially whether this tool could be more useful then “Rewrite all the readme based on new changes. Make no mistakes”, for instance, in environments where AI usage is restricted