Six-skill presentation system: ingest talks into a rhetoric vault, run interactive clarification, generate a speaker profile, create presentations that match your documented patterns, produce the deck illustrations + thumbnail visual layer, and publish talk pages to a Jekyll shownotes site. Includes a 111-entry Presentation Patterns taxonomy (81 observable: 62 patterns + 19 antipatterns; 30 unobservable: 21 patterns + 9 antipatterns) for scoring, brainstorming, and go-live preparation.
—
—
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
View guide
Low
Low-risk findings worth noting
Reading the room's nods as evidence that the material landed. The content is good, the delivery is smooth, everyone is agreeing — and nothing is being retained, because agreement and recognition are not learning, and a nodding room is exactly what a forgetting room looks like from the stage.
This is the hardest antipattern in the catalog to see, because from the stage it is indistinguishable from success. The room is attentive. Heads move. People laugh in the right places. Nobody is on their phone. At the end there is warm applause and three people come up to say it was great. Every signal a speaker has access to says the talk worked.
The signals are measuring the wrong thing. Nodding indicates recognition — "yes, that parses, that follows, I could have told you that" — and recognition is precisely the sensation that fluent, well-constructed explanation produces. It is the fluency illusion, and its cruelest property is that it scales with the quality of your delivery. The better your talk, the smoother your explanation, the cleaner your slides, the more confidently the room will nod, and the more certain everyone in it will be that they have learned something. Roediger and Karpicke's learners predicted that rereading would serve them better than testing; it served them worse. The room does not know what it knows, and it is reporting its confidence to you in good faith.
The failure is not the nodding. The failure is the absence of any moment in the talk where the audience had to produce something. Watch for it structurally: across forty-five minutes, was there a single beat where the room supplied an answer, a prediction, a number, a guess — anything at all — rather than receiving one? In most talks, including most very good talks, the honest answer is no. The audience's entire job was to sit and agree, and they did it well, and they will remember the one joke about the printer.
This is what separates The Nodding Room from _anti_lipstick-on-a-pig.md. Lipstick is bad content behind good visuals, and the audience is being cheated. Here the content is genuinely good, the speaker is genuinely competent, and everyone is behaving honorably. The talk is simply built as a broadcast, and broadcasts are received rather than learned. That is why this antipattern survives so well: nothing in it feels like a mistake, and every incentive the conference circuit offers — the ratings, the applause, the hallway compliments — rewards it.
The speaker-side twin makes it self-sealing. Feedback arrives as satisfaction scores and hallway praise, both of which measure how the talk felt, and neither of which has much to do with what survived the week. A speaker optimizing on those signals will make the talk smoother, more fluent, and more agreeable every year — which is to say, they will make it nod harder, and they will be optimizing directly away from retention while every number goes up. There is no correction available from inside that loop, which is the only reason this file exists.
carnegie-hall.md, "Rehearse by Retrieval, Not Rereading"). Same illusion, one seat over.The cheapest available fix is one genuine retrieval-beat: at one section boundary, ask instead of remind, and wait three seconds. The pause will feel much longer to you than to them.
The second is one guess-first before the talk's most surprising reveal — make the room commit to a number before you show the real one.
The third is a diagnostic rather than a fix, and it is the one worth internalizing: ask what the room could reconstruct tomorrow, not what they agreed with today. If the honest answer is "the printer joke," the talk is a Nodding Room regardless of the ratings.
Then stop trusting the instruments. Nods, laughs, and satisfaction cards measure whether the talk was pleasant. If you want to know whether anything stuck, you have to ask a question and listen to the answer — in the room via retrieval-beat, or two weeks later via spaced-followup. Those are the only two instruments in this catalog that measure retention rather than mood, and both work by making somebody produce something.
The vault detects this structurally, not from audience reaction — retention is invisible in a recording, but its absence has a clean signature:
retrieval-beat or guess-first across the entire talk — no prompt at any point requires the audience to produce an answerDo not flag a talk merely for being tell-shaped. Keynotes, inspirational talks, and performance-shaped talks legitimately contract with the audience to be watched, and this antipattern does not apply to them. The flag requires a talk whose stated purpose is teaching — a tutorial, a deep dive, a workshop, a "how we did X and how you can too" — combined with the structural signature above. A talk that never intended to be learned from cannot fail to be learned from.
Use strong_evaluable_from, evidence_requirements, and not_evaluable_when above to evaluate positive evidence.
Current catalog artifacts may support positive detection only. Because absence_evaluable_from is null, no delivery video, transcript, rendered or native deck, comparison artifact, or claim of full coverage authorizes an absence finding; when no positive signal is established, record not_evaluable, not absent.
Relates to Dimension 4 (Audience Interaction) as the primary axis — the antipattern is the null case of audience interaction, dressed as a room that is engaged. Relates to Dimension 12 (Pacing Clues), since a talk with no production beats has no gather-and-consolidate rhythm and runs as an undifferentiated stream. Relates to Dimension 14 (Areas for Improvement), where it belongs in a distinct category: unlike most entries in Dimension 14, this one flags talks that are already good and would be improved by the one thing their positive feedback will never ask for.
The direct inverse is retrieval-beat — the antipattern is definitionally the absence of it, and one instance of the pattern is the minimum viable refutation. guess-first is the other half of the antidote, covering material the audience has not yet been given.
It is adjacent to but distinct from _anti_lipstick-on-a-pig.md (bad content, good visuals — here the content is good), from _anti_flyover.md (contempt for the audience — here the speaker respects them and serves them attentively), and from _anti_disowning-your-topic.md (misreading the room as disengaged — this is misreading the room as learning, and is in some sense its mirror image: both are errors in translating audience reaction into truth).
It has a feedback-loop relationship with red-yellow-green.md, whose satisfaction cards will reliably score a Nodding Room green (see that file's "Smile Sheets Do Not Measure Retention"), and with crucible.md, whose iterate-on-feedback engine will faithfully optimize the talk deeper into the antipattern if the feedback it consumes is satisfaction rather than recall.
.tessl-plugin
rules
skills
illustrations
presentation-creator
references
patterns
build
deliver
prepare
scripts
shownotes-publisher
vault-clarification
vault-ingress
references
scripts
vault-profile