← ClaudeAtlas

awb-stress-testlisted

WHAT: pressure-test a non-trivial change before you build it — run it past a fixed set of independent lenses so blind spots surface on paper, not in production. Two modes: design stress-test (expert viewpoints judge whether the approach is sound, returning a GO / CAUTION / STOP verdict) and edge-case decomposition (failure-axis lenses enumerate what could break, producing a severity-ranked list that feeds tests). USE WHEN: about to implement or test something non-trivial and you want the risks and edge cases first ("stress test this", "what could go wrong", "edge cases", "is this design sound", "second opinion before I build", "poke holes in this plan"). DO NOT TRIGGER: a one-line / low-risk change; choosing between approaches (that is a research skill); reviewing code already written (that is a review skill); a known bug with a clear cause (that is a debug skill).
doivamong/agent-workbench · ★ 2 · AI & Automation · score 78
Install: claude install-skill doivamong/agent-workbench
# Stress-test a change before you build it > **Announce on activation:** "Using awb-stress-test — I'll run this past fixed lenses to surface risks and edge cases before any code." The cheap-to-fix mistakes are the ones a single pass over your own plan never sees: you read the design the way you intended it, so the blind spot stays blind. This skill imposes a *fixed set of independent lenses* — the same checklist every time — so the same gaps get caught every time, on paper, before they cost a rebuild or a production incident. It runs **before** code exists; it does not review code already written. ## Scope - **Does:** pressure-test a *proposed* change through fixed lenses — either to judge whether the design is sound (mode A) or to enumerate how it could break (mode B) — and synthesize the result into a decision or a test list. - **Does NOT:** compare alternative approaches (that is a research skill — this assumes the approach is already chosen), review or refactor written code (that is a review skill), or diagnose a bug that already happened (that is a debug skill). It surfaces candidates; it does not prove them — evidence does that (see *Honest limit*). ## Pick the mode | Mode | Question | Lenses | Output | |---|---|---|---| | **A — Design stress-test** | "Is this approach sound enough to build?" | expert viewpoints | a GO / CAUTION / STOP verdict + the conflicts behind it | | **B — Edge-case decomposition** | "What could break once it's built?" | failure ax