commit 0705dc84046e1a5cafa01f647254e2e2014d8e00
parent d1cf5afe5eb7d362485444da55fc6d78a2e4145a
Author: I Mean I'm Just Saying <imeanimjustsaying@kiwifarms.st>
Date: Mon, 5 Oct 2026 03:34:27 -0400
records: PUBLISH.md on composing reports and the cited scope; [Unreleased] bullets
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Diffstat:
3 files changed, 29 insertions(+), 1 deletion(-)
diff --git a/PUBLISH.md b/PUBLISH.md
@@ -46,7 +46,7 @@ entry points in `common/publish/build.ts`.
| To | Editor | `pnpm ops` | `pnpm archilyzer …` |
|---|---|---|---|
-| Build one site | a site's **Publish** tab → *Build static export* | `build-site` | `build site <id> [--nodata] [--skip-archives]` |
+| Build one site | a site's **Publish** tab → *Build static export* | `build-site` | `build site <id> [--nodata] [--skip-archives] [--allow-missing-media]` |
| Deploy the built site | *Deploy to production* / *Deploy preview* | `deploy-site` | `deploy site <id> [--preview <branch>]` |
| Build, then deploy | *Build & deploy* | `build-deploy` | `build site <id>` then `deploy site <id>` |
| Build every site | /sites → **Build all sites** | — | `build all [--skip-archives]` |
@@ -74,6 +74,32 @@ is not on disk, a clip over 24 MiB or an invalid report is listed and fails the
(exit 1, or a failed job) — fetch the window or persist the video, capture the
post, and run it again; what is already cut is reused.
+The build's compose then writes the reports from what prepare left
+(`common/publish/composeReports.ts`): each report's page, its citations as
+`citations.json` and `citations.csv`, its cited stills (never a saved source copy),
+one page per cited moment with the transcript lines around it, and the prepared clips
+and captures — only those the published reports cite. Every quote is checked as it is
+composed: a span's against its cues within 5 s either side (a record whose `en` track
+has no cues is read from `en-orig`), a post's against its text, scored as the share of
+the quote's words found there; the score, the time and the method are written into the
+citation, replacing any typed by hand. Compose fails, before it writes any of it, with
+the list of everything wrong: an invalid report, a citation of a channel outside the
+site or a post the site may not carry, a missing record, still or post, a quote below
+60 %, and a citation whose media was not prepared, or was cut for a span the report no
+longer cites. `--allow-missing-media` (on `compose site` and `build site`) lets the
+last two through; their pages render without a clip.
+
+A **cited** site (`publish: "cited"`) publishes its reports and nothing else. Its
+compose removes every corpus-shaped file the shared `export/public` holds — summaries,
+stats, transcripts, subs, posts, digests, archives, the duplicates, tags, aliases and
+chart files, the service worker — writes `site.json` with no channels and `corpus.json`
+with `site.scope: "cited"` (spec 5), and an `llms.txt` and sitemap listing the reports
+and moment pages. After `next build`, the built `out/` is audited: a cited build that
+holds anything but `_next/`, the reports, the moment pages, the cited media, the
+shell's own pages and files, or a file over 25 MiB, or more than 20,000 files, fails
+the build, and every deploy path refuses it. A site switched to cited is refused at
+deploy until it is built again.
+
Deploy-only ships whatever is in `export/out`, which the basic build composes one
site at a time into a single shared directory — so it **refuses, before starting a
job, if `export/out` holds a build of another site** (or no build at all), naming the
diff --git a/editor/CHANGELOG.md b/editor/CHANGELOG.md
@@ -1,6 +1,7 @@
# Changelog
## [Unreleased]
+- **A site's build composes its reports, and a site that publishes only its reports ships nothing else.** Every site's compose now writes the reports its `site.json` publishes: each report's page and its citations as `citations.json` and `citations.csv` under `/reports/<id>/`, its cited stills, a page per cited moment with the record, the transcript lines around the span and every report that cites it, and the clips and post captures `archilyzer reports prepare` made for it, only the cited ones. Each quote is checked against the record as it is composed (a span's against its cues within 5 s either side, read from `en-orig` when the `en` track has no cues; a post's against its text) and the score, time and method are written into the citation, replacing any typed by hand. The build stops with the list of every problem before anything is written: an invalid report, a citation of a channel outside the site or of a post the site may not carry, a missing record, still or post, a quote that matches less than 60 % of what the record says, and a citation without prepared media or with media cut for another span (`--allow-missing-media` on `archilyzer compose site` and `build site` lets those two through, without a clip). A site with `publish: "cited"` removes everything corpus-shaped from `export/public` before it writes its reports, and its built `out/` is checked against what a cited site may hold: anything else, a file over 25 MiB or more than 20,000 files fails the build, and every deploy path (the Publish tab, `deploy site`, Build & deploy, Build & deploy all, the container build) refuses it, as it refuses a site set to cited whose last build was a full one. The hub's compose removes a report site's files too.
- **A site has a Reports tab.** `/sites/<site>/reports` lists every report under the site's `reports/` directory — the published ones in their order, then the drafts — with its kind, dates, sections, claims, citations by kind and, for a fact-check, how many claims carry each verdict. Each report's problems, from the same checker the prepare step and the build use, open under it. A draft with no problems can be published, and a published report moved up or down or unpublished; each writes only the site's `reports` list, applied to the list as it is on disk at that moment, so it never overwrites another change to the site. "Prepare evidence media" queues the `reports-prepare` job, and beside it the tab shows the last prepared media (moments by kind, total size, problems by kind) and links the last prepare job. What the site publishes (full or cited) is shown with a link to Settings, where it is changed.
- **A site can say what it publishes, and which reports.** `site.json` takes `publish` — `"full"`, the searchable corpus every site has been (the default, never written), or `"cited"`, only the site's reports and the moments they cite — and `reports`, the ordered ids of its published reports (each a slug; invalid and repeated ids are dropped). The site form has a Publish control and lists the site's reports read-only; saving the form keeps the stored list. A cited site still builds as a full one until the reports pipeline applies the scope. SITE.md documents both keys.
- **A report and its citations now have one written format, checked before anything is built from them.** A cited report is a `report.json` (`archilyzer-report`, version 1): a summary, then sections of claims, each claim with an optional verdict, the reviewed document's own sentence, findings in markdown and the citations it rests on. A citation is one of five kinds — a span of a video, a span of an audio record, a post, a sentence of a source document, or a web page — with a verbatim quote, and is cited from any markdown in the report as `[label](cite:<id>)`. The checker lists every problem at once with where it is: a citation, a source or a `cite:` link that names nothing, a span that ends before it starts or runs past 120 seconds with its context, a still that points outside the report's folder, an id used twice. Each cited span and post has one page address, `/m/<channel>/<id>/<start>-<end>/` or `/m/<channel>/<id>/`. Nothing builds or shows reports yet. The fact-check verdicts (Corroborated, Partly true, Contradicted, Not found, Untestable) and their colours are now kept in one place, which the report video's stamps and tally read too. `REPORT.md` and `CITATIONS.md` list every key.
diff --git a/export/CHANGELOG.md b/export/CHANGELOG.md
@@ -1,6 +1,7 @@
# Changelog
## [Unreleased]
+- **`corpus.json` is spec 5: it names a site's reports, and a site that publishes only reports says so.** A site with reports adds `reports` to its `corpus.json` (`index`: `/reports/index.json`, the count, and how to read a report's page, its citations and its moment pages) and a Reports section to `llms.txt`; its sitemap lists the report and moment pages. A site that publishes only its reports has `"scope": "cited"` and its audience under `site`, no channels and zero totals, an `llms.txt` that lists its reports and how their citations and moment pages are read, and a `site.json` with no channels. A reader that does not know spec 5 sees an empty corpus there. Needs a rebuild and deploy of each site.
- **A site can show cited reports, and every citation opens on a page of its own.** A site built with reports has a **Reports** link in its header and a page at `/reports/` listing them. A report's page has its title, subtitle, dates and the document under review with its archive links; a fact-check's tally of verdicts; the summary; the sections and their claims, each with its verdict, the document's own sentence as an image, the findings and the evidence cards; a numbered reference list; and links to download its citations as JSON and CSV. A citation in the text shows as its words plus a number: hovering it, focusing the number or tapping it once shows a card of the citation (the quote, who said it and when, a picture or the post's screenshot, and how closely the quote matched the transcript when it was checked); the words open what it cites and the number jumps to its reference. A cited span of a video or audio record opens at `/m/<channel>/<id>/<start>-<end>/`: a short clip of the span with a little context either side, the quote, the transcript lines around it, the record's title, channel and date, a link to the original at that time, and every report on the site that cites it. A cited post opens at `/m/<channel>/<id>/` with its screenshot and text. A site that publishes only its reports (`site.json` `publish: "cited"`) opens on the report index and has no search, Ask AI, downloads or duplicates. A site with no reports is unchanged. Needs a rebuild and deploy of each site.
## [0.11.1] - 2026-10-01
