TFactory — a guided tour

TFactory generates and runs tests across modality lanes (unit, browser, API, integration, mutation) and grades every generated test on a 5-signal verdict — coverage delta, stability re-runs, mutation kills, lint, and semantic relevance — then posts a ranked triage report to your PR. Meaningful tests, not a green bar.

TFactory unit lane — generate, run, grade — The unit lane in motion — generate, run, and grade tests on the 5-signal verdict

Tests you can trust

Each modality lane produces real evidence — screenshots, video, HAR, mutants — and every generated test is scored, so what lands on your PR is a ranked, explained verdict rather than a coverage number.

TFactory polyglot run — **Polyglot**One spec, many languages — the same lane across stacks.

TFactory API lane — **API lane**Generated API tests with real HAR evidence.

TFactory graded triage report — **Verdict**A graded, ranked triage report — the 5-signal score, not coverage theatre.

Feeds the live verify map, and proves the verdict

TFactory exposes per-subtask lane and started_at / completed_at timing on /api/tasks/{id}, which CFactory aggregates into the test-stage of its live execution diagram — a lane pipeline (unit → browser → api → integration → mutation) you watch fill in live, with a subtask stuck driving the stalled state. The verify leg also honours the RFC-0001a evidence gates: reaching a terminal phase is not a pass — a verify only claims success with a non-null verdict and tests actually executed, so a null-verdict run is reported as failed.

← Back to the products