Source Quote Evidence
Da «qualcuno ha detto» a «ecco chi, ecco dove, ecco la prova».
Trasforma un'affermazione citata in un momento verificabile sullo schermo: la citazione compare parola per parola, le prove a supporto si evidenziano sul posto, e la citazione della fonte resta sullo schermo abbastanza a lungo da poterla controllare.

Installala in 3 passaggi
Nessuna registrazione. Nessuno strumento da riga di comando. Scegli la tua AI, copia una riga e incollala.
Preferisci fare da te? Acquista la skill ($4.99), scarica lo ZIP dalla conferma d'acquisto, decomprimilo nella cartella mostrata sopra e riavvia la sessione della tua AI — stesso risultato. L'hai già acquistata? Il tuo assistente AI può scaricarla per te.
Cosa contiene
Un file SKILL.md (la guida operativa) + un modello di partenza. Un'anteprima:
Source Quote Evidence
When to use
- Your video quotes a study, report, interview, or public statement and you want viewers to trust the quote instead of wondering whether you paraphrased it into something stronger
- You're making an explainer that cites several sources and need one consistent treatment so every claim carries the same weight
- Your script says "research shows" — which is exactly the phrase that makes skeptical viewers stop trusting you
- You're rebutting or confirming a claim and the audience needs to see the original wording, not your summary of it
- You quote the same source more than once and want viewers to recognize it on sight the second time
Workflow
- Capture the quote verbatim (input: source document → output: exact quote + citation)
Copy the quote character for character from the original source — no silent tidying, no "basically." If you must trim, mark every cut with an ellipsis (…) and never trim in a way that changes the meaning. Record the full citation now: author or speaker, publication or venue, date, and page/paragraph/timestamp. If you can't find the original source, don't use the quote — a quote traced only to a screenshot of a screenshot is not evidence, it's rumor with formatting. - Pick the evidence span (input: verbatim quote → output: one highlighted span)
Inside the quote, choose the single span that carries the claim — the number, the verdict, the date — usually 5–15 words. This span gets the highlight treatment later; everything else stays plain. If the quote needs two highlights to make sense, it probably needs two quotes — split it. The highlight answers the viewer's question "which part am I supposed to believe?" before they ask it. - Stage the quote card (input: quote + citation → output: quote card layout)
Build the card: large quotation mark, quote text at a size readable on a phone (minimum 36 px at 1080p for the quote body), citation in smaller text beneath (author, source, date), and a thin rule separating quote from citation. Keep the card to one idea and under 40 words — longer quotes get trimmed with ellipses or split across two cards. The citation is part of the card, not a footnote: it stays on screen for the entire time the quote is up. - Reveal word by word (input: quote card → output: 0.0–3.0s reveal animation)
Reveal the quote in phrase chunks (3–6 words per chunk), each chunk fading + rising 8 px over 0.25s, timed to the voiceover reading it — the chunk the voice is speaking is the chunk appearing. The evidence span reveals last, with a highlight sweep: a soft silver bar slides behind the span over 0.4s and stays. Never reveal the whole quote at once while the voice is still on the first sentence — viewers will read ahead, stop listening, and the reveal will feel redundant. Pacing target: the final word lands within 0.5s of the voice finishing the quote. - Show the evidence in place (input: revealed quote → output: 1–2s evidence hold)
After the reveal, hold the full card for 1–2s with the evidence span highlighted and the citation visible, and — where the source allows — flash the source context behind the card: a faded page, a document header, a timestamped video frame, at 15–25% opacity. This is the "you can check this" beat. If the source is a video, show the original frame; if it's a document, show the page header with the date. Never fake source context: a stock photo of "a document" behind a real quote is worse than a plain background. - Tag the source for recall (input: finished card → output: source tag)
Give each distinct source a small persistent tag — a two-letter code or a short name chip ("JAMA 2024", "Senate hearing") — that appears in the card's corner and recurs whenever that source is quoted again. On the second use, viewers recognize the tag before reading the citation. Keep a source list at the end of the video (or in the description) with a link or full reference per tag, so the tag is a pointer, not a dead end.
Anteprima del modello
Source Quote Evidence — Quote Card Template
One card = one quote = one claim. Fill in every field before animating.
Quote card #{n}
Claim this card proves: {one sentence — what the viewer should believe after seeing it}
Verbatim quote
> {exact quote, ≤40 words, trims marked with …}
Evidence span (5–15 words, gets the highlight): "{span}" Words trimmed: {list each cut, or "none"}
Citation (stays on card the whole time)
- Author / speaker: {name}
- Publication / venue: {name}
- Date: {date}
- Page / paragraph / timestamp: {location}
- Access note: {open / paywalled — abstract cited / offline}
Source context (real artifact only)
Il pacchetto della skill (SKILL.md + modello) è scritto in inglese.


