Handbook

What is Forge STI?

Forge STI (Software Testing Intelligence) is a private Forge product for governed test strategy, plans, cases, and campaigns. It treats quality work as structured, reviewable artifacts — not chat-only test ideas.

Updated

Core thesis

Many teams use AI to generate pytest or Playwright snippets. STI goes further:

  1. Layered abstractions — strategy → plan → suite → case → execution binding → campaign flow (see Layered test artifacts).
  2. lmeta orchestration — multi-step campaigns (generate → evaluate → run → evidence) as reviewable .lmeta documents.
  3. LCDL for judgment — when case design, triage, or coverage analysis affects a release decision, route through governed task contracts.
  4. Cursor as control surface — rules, composer prompts, and skills author STI artifacts interactively.
  5. ForgeRun evidence — test run summaries attach to the spine; Lenses remains system of record for review.

What STI is not

  • Not a TestRail clone or generic test-management SaaS.
  • Not a replacement for Blueprints testing methodology — STI executes ISTQB-aligned practice under Forge governance.
  • Not a second run spine — visibility and approval stay in Lenses.

Maturity

This handbook is defined (architecture and schemas). Runnable LCDL tasks, lmeta campaign flows, and Fleet templates are planned follow-ons.

See Platform composition for boundaries vs Platform, LCDL, Fleet, Lenses, and Blueprints.