diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md
index 7e4f604f..3f7210fe 100644
--- a/CONTRIBUTING.md
+++ b/CONTRIBUTING.md
@@ -26,8 +26,52 @@ Exercises can be tested against the test runner using
 bin/verify-exercises-in-docker
 ```
 
+## Formatting and linting
+
+CI runs the following checks on every pull request, so please run them locally before submitting.
+
+All Zig code — exercise stubs, example solutions and test files — must be formatted with `zig fmt`:
+
+```bash
+zig fmt --check .
+```
+
+The remaining checks use [configlet], Exercism's track maintenance tool.
+Fetch it once before running any `bin/configlet` command — this downloads the binary to `bin/configlet`:
+
+```sh
+bin/fetch-configlet
+```
+
+Or for Windows:
+
+```sh
+bin/fetch-configlet.ps1
+```
+
+The track configuration must pass configlet's linter, and the JSON files must be formatted with configlet's formatter:
+
+```bash
+bin/configlet lint
+bin/configlet fmt --update
+```
+
+When adding an exercise or an approach, generate any new UUIDs with configlet:
+
+```bash
+bin/configlet uuid
+```
+
+Practice-exercise test files are generated from the canonical data in [`problem-specifications`][problem-specifications]; edit the generator in `generators/exercises/` and regenerate:
+
+```bash
+bin/generate <slug>
+```
+
 [guidelines]: https://exercism.org/blog/contribution-guidelines-nov-2023
 [zig-forum]: https://forum.exercism.org/c/programming/zig/199
 [building-tracks]: https://exercism.org/docs/building/tracks
 [style]: https://exercism.org/docs/building/markdown/style-guide
 [unnecessary-test-runs]: https://exercism.org/docs/building/tracks#h-avoiding-triggering-unnecessary-test-runs
+[configlet]: https://github.com/exercism/configlet
+[problem-specifications]: https://github.com/exercism/problem-specifications