Users and Use Cases¶
Summary¶
Meridian serves multiple engineering roles, but they do not use the product in the same way. Good documentation should make those paths explicit.
Platform and observability engineers¶
This is Meridian's primary audience.
Typical use cases:
- editing shared Collector configs in Git
- verifying route, processor, or exporter changes locally
- producing artifacts that other reviewers can inspect
- capturing the difference between expected negative behavior and an actual regression
These users usually care most about validate, check, diff, and artifact quality.
Reviewers¶
Reviewers often do not run the CLI themselves. They consume the output from CI or from an artifact bundle attached to a pull request.
They care about:
- clear risk highlights
- graph artifacts
- semantic validation stage results
- runtime evidence and contract results
- summaries that explain failure without requiring reruns
CI maintainers¶
CI maintainers care about consistency and scriptability.
They need:
- predictable exit codes
- machine-readable JSON output
- PR-comment and step-summary support
- durable artifact uploads
- enough context to debug failures from the workflow logs and bundle
Meridian maintainers¶
Maintainers need deeper confidence in Meridian itself. That is why the repository includes a repo-owned k3s fixture and a dedicated regression workflow outside the normal user path.