Try to kill the artifact, not improve it — return the one root flaw and the cheapest shot that fells it.

Goal

Where shower asks "can a stranger follow this?" (comprehension), redteam asks "does this survive someone trying to break it?" (validity). It is the superset reflex: where factchk checks one fact and mandela audits one validation, redteam attacks the whole plan and collapses it to a single root. The highest-value output is never a checklist — it's one root cause plus the cheaper experiment hiding inside the elaborate plan.

Workflow

Pin the load-bearing assumption(s) — what must hold for the whole thing to stand.
Attack on whatever axes apply — the general ones first: a load-bearing fact that may be false (run factchk), confabulation (a post-hoc story treated as ground truth), analogy-mistaken-for-isomorphism (structure assumed to transfer across domains where it doesn't), future-tense suture (the argument leans on a result that doesn't exist yet), and the sharpest — cites a principle but implements its opposite; and for empirical / research plans, leakage (run mandela) and statistical power / family-wise α (a true hypothesis auto-failing from uncorrected tests; a near-zero-power condition that rubber-stamps regardless of truth).
Collapse the findings to the single root cause — the one whose failure makes the others moot.
Find the first nail — the cheapest falsification of the load-bearing assumption: the check (in time / cost / sample) that could kill it before the expensive program runs.
Return { root, first_nail } — not a list.

Rules

Kill, don't improve — improving is a different reflex.
Compose, don't duplicate — invoke factchk and mandela for their axes; you own the synthesis to one root.