Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →platformplatform/test
Run backend (.NET) xUnit unit and integration tests via the developer CLI.
Passed 18 of 20 test cases with the skill, and 4 of 20 without it.
and 13 more in the full report
16 fixed4 passed either way2 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 16 test cases the skill fixed. Same model, same prompt, with and without the skill.
Run the full test suite via the developer CLI, but make sure to supply /home/runner/work/repo/developer-cli as the project path parameter so dotnet locates it.
| Checked | Without skill | With skill |
|---|---|---|
| Uses developer-cli as project argument value | Passed | Passed |
| Does not use absolute path in project argument | Missed | Passed |
The instructions your model reads when the skill loads.
dotnet run --project developer-cli -- test [--self-contained-system <name>] [--filter <expr>] [--no-build] [--exclude-category <cat>] --quiet
Use developer-cli exactly as written - do not expand to an absolute worktree path.
Backend only - there is no frontend test runner.
--self-contained-system <name> - narrows to one SCS (e.g. account, main)--filter <expr> - forwarded to dotnet test --filter to scope to a subset of tests--no-build - skip rebuild before running (faster after a recent build)--exclude-category <cat> - defaults to Noisy; pass an empty string to include themNo arguments runs every test across every SCS.
After build succeeds, run format, lint, test in parallel with --no-build.
dotnet run --project developer-cli -- test --quiet # all tests
dotnet run --project developer-cli -- test --self-contained-system account --quiet # one SCS
dotnet run --project developer-cli -- test --filter "FullyQualifiedName~LoginTests" --quiet # filter by name
dotnet run --project developer-cli -- test --no-build --quiet # after a recent build
Verbose output goes to a log file. On success the CLI prints a one-line summary (totals + duration); on failure it lists the failed test names plus the log path - read the log if you need stack traces. Without --quiet every test result streams into the conversation.
test cases passed, vs 4 / 20 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.