scientific-methodlisted

Use this skill whenever someone doubts a number, demands rigorous proof of a cause, or wants something invented and proven — any time a benchmark, metric, ceiling, or root-cause story must be verified rather than trusted, or a limit must be broken with a built, measured mechanism. Typical situations: a measurement looks suspicious ("2x faster but I don't trust it"); an incident needs its true cause before a decision ("the team blames X — confirm it, no guessing"); ANY production incident, outage, brownout, latency/error spike, regression, or "started failing" / "load suddenly went 5x" / "every component met spec but it still broke" report where the cause is not yet proven — especially when a recent deploy or change is suspected ("is the deploy to blame", "this coincided with the rollout"), since the obvious suspect is exactly what needs a control case; an asserted ceiling or "impossible" limit needs breaking or proving real; numbers need validation before publishing; a problem needs an invented solution that
88plug/scientific-method · ★ 1 · AI & Automation · score 57

Install: claude install-skill 88plug/scientific-method

# Scientific Method A working method for discoveries, distilled from real campaigns: a GPU codec that broke four "physical" performance walls because someone asked *"are we sure this ceiling is right?"*, a fleet investigation that killed two plausible-but-wrong root causes with control cases before blaming a vendor, and a research fan-out that burned five agents arguing about a question one HTTP GET would have settled. The core stance: **every asserted limit, cause, or claim is a falsifiable hypothesis until measured.** You do not get to say "wall", "impossible", "root cause", or "confirmed" without a probe behind it. And the method's purpose is not only to verify — **it is to invent**. A confirmed limit is not an ending; it is the entry criterion for an invention campaign: design mechanisms past it (Pattern B + the TRIZ separations), build them, measure them against tuned baselines or pre-set acceptance criteria, and close every claim with a provenance search. The same falsification discipline that kills bad explanations is what makes an invention claim *provable*: demonstrated under re-run, or honestly proven impossible — never asserted. The full invention pipeline lives in [references/invent.md](references/invent.md); for world-record attempts on certificate-verifiable open problems, the campaign playbook is [references/open-records.md](references/open-records.md). ## The loop ### 0. Cheapest falsification first Before designing anything, ask: what is the single chea