Answers and receipts

A receipt is one completed answer run: the auditable record of what a single engine said to a single prompt on a single date, with its ranked brand mentions, sentiment evidence, citations, and any ad or shopping surface. Every metric in MentionFlow traces back to one of these.

What it does

Answers is the ground truth beneath every chart. Where a metric is a number, a receipt is the answer that produced it — the verbatim text an engine returned, plus everything MentionFlow extracted from it. If you ever want to know why a number moved, you open the receipts behind it.

How to use it

• Scan the feed. The feed shows up to 500 rows with engine and type chips, a presence summary, and a searchable table. Click the chevron on any row to expand a preview of the full answer inline — the text is fetched on demand, so you can skim what several engines actually said without opening each one. For the complete record, follow Open receipt.
• Trust the metrics boundary. Incomplete runs (queued or failed) appear in a separate strip and never enter metrics.
• Fact-check across engines. An optional knowledge-base fact-check card compares up to 8 answers — one per engine — against your knowledge base.
• Open a receipt. Click a row for /answers/[runId]: the highlighted answer text, "Brand ranking in this answer" (every mention in order, including untracked ones, each with a Track button), the citations, sponsored-ad and shopping cards, and per-engine mini-stats.

Reading a receipt

A receipt is where an answer comes alive. Open one from any feed row — or from a prompt's detail page — to inspect exactly what the engine said and what MentionFlow read from it. It is the fastest way to explain why a number is what it is.

• Hover a brand mention. On a receipt (/answers/[runId]), the brand names in the answer are highlighted. Hover one for a card showing its position (for example "Mention #2 of 5"), its sentiment — or unscored when the run wasn't scored — whether it is you or untracked, and the verbatim quote the sentiment was read from.
• Follow a citation. Where the answer links to a source, the link carries that site's favicon. Hover it for a card with the domain, its source classification, the page title, and a Source intel link through to that domain under Sources. A +n marker means the engine folded that many more sources behind it — the full list is in the citations panel. (Citation cards appear on both the receipt and the prompt detail page.)
• See what the sentiment was read from. Positive and negative mentions shade the exact passage the score came from — teal for positive, rust for negative — so you can audit any score down to the sentence. Neutral and unscored mentions are left unshaded, and a short legend ("read as positive" / "read as negative") appears whenever a passage is shaded.

How it's computed

Extraction runs in two passes:

1. A deterministic pass decides presence and order — who is mentioned, and in what sequence.
2. A language-model pass (temperature 0, strict JSON) produces the canonical cross-brand mention order plus sentiment and an evidence quote.

Two more rules keep receipts honest:

• Ambiguous brand names require domain-form evidence before they count, so a common word that happens to match a brand name doesn't inflate presence.
• Surface stats omit any flag with zero rows rather than printing 0% — an absent ad surface simply isn't listed.

Receipts age out with the data horizon: a receipt older than 56 days returns a 404.

Limits

• The feed is capped at 500 rows.
• Fact-check needs a knowledge base, edit rights, and an OpenAI key.
• CSV export is available at /api/export/answers.

Related

• Prompts — the prompt each receipt answers.
• Sources — the citations a receipt carries, aggregated.
• Overview metrics — the metrics receipts feed.
• API / MCP: GET /api/v1/answers (REST cap 200 rows) and the list_answers MCP tool (cap 50 rows).