Auto-diagnose test gaps from file changes and track test outcomes
Comparison
Fix-test gives you two ways to engage tests: let
TestMaintenanceWorkflow plan the work from file changes, or call
the test_runner functions to measure tests directly. They answer
different questions and often pair.
| Capability | TestMaintenanceWorkflow (planning) |
test_runner functions (measurement) |
|---|---|---|
| Question answered | "What test work does this change imply?" | "What did the tests actually do?" |
| Import | from attune.workflows import TestMaintenanceWorkflow |
from attune.workflows.test_runner import run_tests_with_tracking |
| Input | File events / project index | A suite name, file list, or coverage.xml |
| Output | A TestMaintenancePlan of TestPlanItems |
TestExecutionRecord / CoverageRecord / FileTestRecord |
| Concurrency | Async (run, on_file_*); sync summary methods |
Synchronous functions |
| Runs tests? | Only in "execute" / "auto" mode |
run_tests_with_tracking / track_file_tests do |
| State | Index-backed; refreshes on change | Persists records to the telemetry store |
| Typical caller | File-watcher, git hook, CI maintenance step | Test runner wrapper, coverage pipeline |
Use the workflow when you want decisions — which files need tests,
how urgent, what is safe to auto-run. Use the test_runner
functions when you want facts — run a suite, capture coverage, look
up a file's status. A common loop is both: run tests with tracking,
then let the workflow plan from the recorded staleness and gaps.