Choose the task you want to finish. You do not need to read every guide in order.
| Start here | Result |
|---|---|
| Install and run the demo | Try a complete original example, then choose a project or global install. |
| Connect a coding agent | Configure MCP, run a connection check, and inspect returned evidence. |
| Investigate your own command | Save repeatable baseline and verification actions in your project. |
| Recheck an existing unit test | Follow one NUnit test through a failure, a patch, and a skipped-test control. |
| Explore original examples | Reduce data-import and asynchronous-update failures and check their fixes. |
| Task | Guide |
|---|---|
| Choose a failure signature or find an option | CLI reference |
| Pass executable arguments without shell parsing | Direct execution |
| Compare a proposed change with a saved baseline | Verify |
| Require evidence that a check actually ran | Execution checkpoints |
| Package and replay selected evidence | Bundles |
| Inspect accumulated evidence storage | Storage inventory |
| Understand the README animation and its controls | Guided demo |
- Result interpretation: CLI outcomes and exit codes, Verify verdicts, and resource limits.
- Compatibility: 1.x contract and migration from 0.x.
- Performance: scope and costs and benchmark method. Measurements describe their workloads and environments.
- Development: contributing, implementation, release procedure, and release validation records.
- Direction: roadmap, adoption priorities, and optional workflow observations.
Install examples use the published version. Features marked Unreleased require a source build and are not available from those package commands yet.