add llm interface

app/api/videos/route.ts

@@ -94,8 +94,22 @@ export async function GET(request: NextRequest) {
 
   const hasMore = offset + videos.length < total;
 
+  const slim = sp.get("slim") === "1";
+  const payload = slim
+    ? videos.map((v) => ({
+        title: v.title,
+        date: v.date,
+        body: v.body,
+        category: v.category,
+        slug: v.slug,
+        duration: v.duration,
+        hasTranscript: v.hasTranscript,
+        jsonUrl: `/json/${v.slug}`,
+      }))
+    : videos;
+
   const response = NextResponse.json({
-    videos,
+    videos: payload,
     total,
     totalIncludingOther,
     hasMore,

app/json/[...meeting]/route.ts

@@ -9,6 +9,12 @@ import {
 import { getCountryName } from "@/lib/country-lookup";
 import { symbolFromSlug } from "@/lib/meeting-slug";
 import { TRANSCRIPT_DISCLAIMER } from "@/lib/config";
+import {
+  buildSpeakerSegments,
+  formatTranscriptAsPlainText,
+  formatSpeakerText,
+  formatTimecode,
+} from "@/lib/transcript-formatting";
 
 export async function GET(
   request: NextRequest,
@@ -92,6 +98,43 @@ export async function GET(
 
     const topics = transcript.content.topics || {};
 
+    const format = request.nextUrl.searchParams.get("format");
+    if (format === "text") {
+      const segments = buildSpeakerSegments(
+        transcript.content.statements,
+        speakerMappings,
+      );
+      const body = formatTranscriptAsPlainText(
+        segments,
+        transcript.content.statements,
+        (idx) => formatSpeakerText(idx, speakerMappings, countryNames),
+        formatTimecode,
+      );
+      const title = video.cleanTitle || video.title;
+      const date = video.date
+        ? new Date(video.date).toLocaleDateString("en-GB", {
+            day: "numeric",
+            month: "long",
+            year: "numeric",
+          })
+        : "";
+      const header = [
+        `UN Transcripts — https://transcripts.un.org/en/${slug}`,
+        [title, video.body, date].filter(Boolean).join(" — "),
+        `Language: ${transcript.language_code}`,
+        TRANSCRIPT_DISCLAIMER,
+        "",
+        "---",
+        "",
+      ].join("\n");
+      return new Response(header + body, {
+        headers: {
+          "Content-Type": "text/plain; charset=utf-8",
+          "Cache-Control": "s-maxage=60, stale-while-revalidate=300",
+        },
+      });
+    }
+
     // Timestamps are already realignment-shifted by the display getter
     // (getTranscriptByKalturaId).
     const transcriptData = transcript.content.statements.map(

app/llms-full.txt/route.ts

@@ -0,0 +1,228 @@
+const DISCLAIMER =
+  "Automatically generated transcript — may contain errors. Not an official United Nations record.";
+
+const CONTENT = `# UN Transcripts — Full API Reference
+
+> Automatically generated transcripts of public United Nations meetings — not official UN records.
+
+UN Transcripts provides searchable, timestamped transcripts of public meetings from UN Web TV (webtv.un.org). Transcripts include speaker identification, topic analysis, and word-level timestamps synchronized to the video. Available in all six official UN languages: English, French, Spanish, Arabic, Chinese, Russian.
+
+Disclaimer: "${DISCLAIMER}"
+
+## Quick start
+
+1. **Search** for meetings: \`GET /api/videos?q={query}&slim=1\`
+2. **Read** a transcript: \`GET /json/{slug}?format=text\`
+
+---
+
+## Search & browse meetings
+
+\`\`\`
+GET /api/videos
+\`\`\`
+
+Returns a paginated list of UN meetings matching the given filters. This is the same endpoint that powers the website's homepage table.
+
+### Query parameters
+
+| Parameter | Type | Description |
+|-----------|------|-------------|
+| \`q\`     | string | Search meeting titles and metadata (not transcript content). Min 2 characters. |
+| \`body\`  | string (multi) | Filter by UN body. Repeat for multiple: \`?body=Security+Council&body=General+Assembly\` |
+| \`category\` | string (multi) | Filter by meeting category. Repeat for multiple. |
+| \`date\`  | YYYY-MM-DD | Filter to a specific date. |
+| \`sort\`  | enum | \`date_desc\` (default), \`date_asc\`, \`title_asc\`, \`title_desc\` |
+| \`offset\`| integer | Pagination offset. Results come in chunks of 100. |
+| \`text\`  | string (multi) | Filter by available documents: \`transcript\`, \`pv\` (verbatim record), \`sr\` (summary record). |
+| \`slim\`  | \`1\` | Compact response — returns only essential fields (recommended for LLM use). |
+
+### Response shape (slim=1)
+
+\`\`\`json
+{
+  "videos": [
+    {
+      "title": "...",
+      "date": "YYYY-MM-DDT00:00:00.000Z",
+      "body": "Security Council",
+      "category": "...",
+      "slug": "sc/{n}",
+      "duration": "HH:MM:SS",
+      "hasTranscript": true,
+      "jsonUrl": "/json/sc/{n}"
+    }
+  ],
+  "total": 42,
+  "totalIncludingOther": 42,
+  "hasMore": true
+}
+\`\`\`
+
+### Notes
+
+- Covers the last 365 days (same window as the website homepage).
+- Without \`slim=1\`, each video object includes ~20 additional fields used by the frontend UI.
+- Use \`hasMore\` + incrementing \`offset\` to paginate through all results.
+
+---
+
+## Read a transcript
+
+### Plain text (recommended for LLMs)
+
+\`\`\`
+GET /json/{slug}?format=text
+\`\`\`
+
+Returns the transcript as plain text with speaker labels. Compact and easy to parse. No time limit — works for any meeting by its slug.
+
+**Optional**: \`?language=fr\` to get a specific language track (default: English if available).
+
+**Example response:**
+
+\`\`\`
+UN Transcripts — https://transcripts.un.org/en/{slug}
+{title} — {body} — {date}
+Language: en
+Automatically generated transcript — may contain errors. Not an official United Nations record.
+
+---
+
+{Country} · {Function} · {Name} [{timestamp}]:
+
+{transcript text...}
+\`\`\`
+
+### Structured JSON
+
+\`\`\`
+GET /json/{slug}
+\`\`\`
+
+Returns full structured data with timestamps, speaker mappings, topics, and word-level timing.
+
+**Optional**: \`?language=fr\` to get a specific language track.
+
+**Response shape:**
+
+\`\`\`json
+{
+  "disclaimer": "Automatically generated transcript — ...",
+  "video": {
+    "id": "...",
+    "kaltura_id": "...",
+    "title": "...",
+    "clean_title": "...",
+    "url": "https://webtv.un.org/en/asset/...",
+    "date": "YYYY-MM-DDT00:00:00.000Z",
+    "duration": "HH:MM:SS",
+    "category": "...",
+    "body": "...",
+    "slug": "..."
+  },
+  "transcript": {
+    "transcript_id": "...",
+    "language": "en",
+    "data": [
+      {
+        "statement_number": 1,
+        "speaker": {
+          "name": "...",
+          "affiliation": "XXX",
+          "affiliation_full": "...",
+          "group": null,
+          "function": "..."
+        },
+        "paragraphs": [
+          {
+            "sentences": [
+              {
+                "text": "...",
+                "start": 12.0,
+                "end": 15.0,
+                "topics": [
+                  {
+                    "key": "...",
+                    "label": "...",
+                    "description": "..."
+                  }
+                ],
+                "words": [
+                  { "text": "...", "start": 12.0, "end": 12.2 }
+                ]
+              }
+            ]
+          }
+        ]
+      }
+    ],
+    "topics": [
+      {
+        "key": "...",
+        "label": "...",
+        "description": "..."
+      }
+    ]
+  }
+}
+\`\`\`
+
+**Key fields:**
+
+- \`data[]\` — speaker turns (statements). Each has \`paragraphs[].sentences[]\` with \`text\`, \`start\`/\`end\` (seconds, float), \`topics\`, and optional \`words[]\`.
+- \`speaker\` — resolved speaker info. \`affiliation\` is ISO 3166-1 alpha-3. \`affiliation_full\` is the expanded country name.
+- \`topics[]\` on each sentence — 0–3 topics this sentence relates to.
+- \`words[]\` — word-level timing (omitted when the provider didn't supply it).
+
+---
+
+## Meeting URL scheme
+
+Slugs are derived from UN document symbols:
+
+| UN Body | Symbol | URL | Example |
+|---------|--------|-----|---------|
+| Security Council | S/PV.{n} | /sc/{n} | /sc/9748 |
+| General Assembly | A/{s}/PV.{n} | /ga/{s}/{n} | /ga/79/21 |
+| GA Emergency Session | A/ES-{s}/PV.{n} | /ga/es{s}/{n} | /ga/es11/23 |
+| GA Committees | A/C.{c}/{s}/SR.{n} | /ga/c{c}/{s}/{n} | /ga/c1/79/7 |
+| Human Rights Council | A/HRC/{s}/SR.{n} | /hrc/{s}/{n} | /hrc/58/59 |
+| ECOSOC | E/{y}/SR.{n} | /ecosoc/{y}/{n} | /ecosoc/2024/10 |
+| Other | — | /meeting/{asset_id} | /meeting/k1tofqtch6 |
+
+Multi-part meetings append \`-part-{n}\`: \`/sc/9748-part-2\`.
+
+---
+
+## Coverage
+
+Public meetings recorded on UN Web TV, including:
+
+- Security Council
+- General Assembly (plenary and all main committees)
+- Human Rights Council
+- Economic and Social Council
+- Other inter-governmental bodies as available
+
+Closed or confidential meetings are not covered (they are not recorded on Web TV).
+
+---
+
+## Known limitations
+
+- **Search scope**: the \`/api/videos?q=\` search covers video titles and metadata, not transcript content. It cannot find meetings based on what was said — only what the meeting is titled/categorized as.
+- **No speaker filtering**: to find what a specific speaker or country said, fetch the full transcript and search within it.
+- **Time window**: search and browse cover the last 365 days, matching the website homepage.
+- **Transcript accuracy**: these are automatic speech recognition outputs, not official records. Names, abbreviations, and document symbols may be misheard. Accuracy varies by speaker and microphone quality.
+- **Languages**: six UN languages are supported (en, fr, es, ar, zh, ru). Not every meeting has transcripts in all languages — it depends on which audio tracks are available.
+`;
+
+export function GET() {
+  return new Response(CONTENT, {
+    headers: {
+      "Content-Type": "text/plain; charset=utf-8",
+      "Cache-Control": "s-maxage=86400, stale-while-revalidate=604800",
+    },
+  });
+}

app/llms.txt/route.ts

@@ -0,0 +1,40 @@
+const CONTENT = `# UN Transcripts
+
+> Automatically generated transcripts of public United Nations meetings — not official UN records.
+
+UN Transcripts provides searchable, timestamped transcripts of public meetings from UN Web TV (webtv.un.org), covering the Security Council, General Assembly, Human Rights Council, ECOSOC, and other inter-governmental bodies. Transcripts include speaker identification, topic analysis, and word-level timestamps synchronized to the video.
+
+Available in all six official UN languages: English, French, Spanish, Arabic, Chinese, Russian.
+
+## How to use (search → read)
+
+- Search meetings: \`GET /api/videos?q={query}&slim=1\` — search meeting titles and metadata with filters for body, category, date, and document type. Paginated (100 per page). Covers the last 365 days. Use \`slim=1\` for compact responses. Note: searches titles/metadata only, not transcript content.
+- Read transcript (text): \`GET /json/{slug}?format=text\` — plain-text transcript with speaker labels, compact for LLM context. No time limit — works for any meeting by slug.
+- Read transcript (JSON): \`GET /json/{slug}\` — structured JSON with timestamps, speakers, topics, and optional word-level timing.
+- [Full API reference](/llms-full.txt): detailed query parameters, response shapes, and known limitations.
+
+## Meeting URL scheme
+
+Meeting pages use human-readable slugs derived from UN document symbols:
+
+- Security Council: \`/sc/{n}\` (e.g. /sc/9748 for S/PV.9748)
+- General Assembly: \`/ga/{session}/{meeting}\` (e.g. /ga/79/21)
+- GA Committees: \`/ga/c{n}/{session}/{meeting}\`
+- Human Rights Council: \`/hrc/{session}/{meeting}\`
+- ECOSOC: \`/ecosoc/{year}/{meeting}\`
+- Other meetings: \`/meeting/{asset_id}\`
+
+## Pages
+
+- [Home](/)
+- [About](/en/about)
+`;
+
+export function GET() {
+  return new Response(CONTENT, {
+    headers: {
+      "Content-Type": "text/plain; charset=utf-8",
+      "Cache-Control": "s-maxage=86400, stale-while-revalidate=604800",
+    },
+  });
+}

components/transcription-panel.tsx

@@ -18,6 +18,7 @@ import {
   formatTimecode,
   formatSpeakerText,
   formatTranscriptAsPlainText,
+  buildSpeakerSegments,
 } from "@/lib/transcript-formatting";
 import {
   TranscriptToolbar,
@@ -269,30 +270,7 @@ export function TranscriptionPanel({
       statementsData: Statement[],
       mappings: SpeakerMapping,
     ): SpeakerSegment[] => {
-      const segs: SpeakerSegment[] = [];
-      if (statementsData.length === 0) return segs;
-
-      let currentSegment: SpeakerSegment | null = null;
-      statementsData.forEach((stmt, index) => {
-        const speakerInfo = mappings[index.toString()];
-        const speakerId = JSON.stringify(speakerInfo || {});
-        const timestamp = stmt.paragraphs[0]?.sentences[0]?.start
-          ? stmt.paragraphs[0].sentences[0].start / 1000
-          : 0;
-
-        if (!currentSegment || currentSegment.speaker !== speakerId) {
-          if (currentSegment) segs.push(currentSegment);
-          currentSegment = {
-            speaker: speakerId,
-            statementIndices: [index],
-            timestamp,
-          };
-        } else {
-          currentSegment.statementIndices.push(index);
-        }
-      });
-      if (currentSegment) segs.push(currentSegment);
-      return segs;
+      return buildSpeakerSegments(statementsData, mappings) as SpeakerSegment[];
     },
     [],
   );