fodda-mcp 1.46.31 → 1.46.97

package/dist/reportBriefing.js

@@ -0,0 +1,662 @@
+/**
+ * 5-Pillar Editorial & Network Synthesis for Report Intelligence.
+ *
+ * Implements the Fodda Editorial Briefing Architecture:
+ * - Pillar 1: Core Tension / Topline Hook
+ * - Pillar 2: Core Shifts & Dramatic Evidence (3-5 structural shifts + metrics)
+ * - Pillar 3: Cross-Graph Network Evidence (validates / contrasts / related with provenance)
+ * - Pillar 4: Human Expert Digital Twin Spotlight (twin position + consult link)
+ * - Pillar 5: Actionable Follow-Up Moves & Quick Links (runnable tool calls)
+ */
+import { callGemini } from './editorialFill.js';
+import { getLiveGraphs, getGraphs, getAnalysts } from './catalogCache.js';
+// ---------------------------------------------------------------------------
+// Evidence Extraction (handles object & array forms from live API)
+// ---------------------------------------------------------------------------
+export function extractEvidenceFromRow(row) {
+    const stats = [];
+    const snippets = [];
+    const caseStudies = [];
+    if (!row)
+        return { stats: [], snippets: [], caseStudies: [] };
+    // 1. Object form: evidence: { statistics: [...], case_studies: [...], quotes: [...] }
+    if (row.evidence && typeof row.evidence === 'object' && !Array.isArray(row.evidence)) {
+        if (Array.isArray(row.evidence.statistics)) {
+            for (const s of row.evidence.statistics) {
+                const statText = s.stat || s.percentage || s.metric || s.number || (typeof s === 'string' ? s : s.text);
+                if (statText)
+                    stats.push(String(statText).trim());
+                const snip = s.snippet || s.description || s.title || (typeof s === 'string' ? s : '');
+                if (snip)
+                    snippets.push(String(snip).trim());
+            }
+        }
+        if (Array.isArray(row.evidence.case_studies)) {
+            for (const cs of row.evidence.case_studies) {
+                const csTitle = cs.title || cs.name || cs.brand || (typeof cs === 'string' ? cs : '');
+                const csSnip = cs.snippet || cs.description || cs.summary || (typeof cs === 'string' ? cs : '');
+                if (csTitle || csSnip) {
+                    const fullCs = [csTitle, csSnip].filter(Boolean).join(': ');
+                    caseStudies.push(fullCs);
+                    snippets.push(fullCs);
+                }
+            }
+        }
+        if (Array.isArray(row.evidence.quotes)) {
+            for (const q of row.evidence.quotes) {
+                const snip = q.snippet || q.text || q.quote || (typeof q === 'string' ? q : '');
+                if (snip)
+                    snippets.push(String(snip).trim());
+            }
+        }
+        if (Array.isArray(row.evidence.items)) {
+            for (const item of row.evidence.items) {
+                const snip = typeof item === 'string' ? item : (item.snippet || item.title || '');
+                if (snip)
+                    snippets.push(String(snip).trim());
+            }
+        }
+    }
+    else if (Array.isArray(row.evidence)) {
+        // Array form: evidence: [{ title, snippet, stat, percentage }]
+        for (const e of row.evidence) {
+            const statText = e.stat || e.percentage || e.metric || e.number;
+            if (statText)
+                stats.push(String(statText).trim());
+            const snip = typeof e === 'string' ? e : (e.snippet || e.title || e.description || '');
+            if (snip)
+                snippets.push(String(snip).trim());
+        }
+    }
+    // 2. Direct top-level fields
+    if (row.stat || row.percentage)
+        stats.push(String(row.stat || row.percentage).trim());
+    if (Array.isArray(row.metrics)) {
+        for (const m of row.metrics)
+            stats.push(String(m).trim());
+    }
+    if (Array.isArray(row.statistics)) {
+        for (const s of row.statistics) {
+            const st = typeof s === 'string' ? s : (s.stat || s.text || s.snippet);
+            if (st)
+                stats.push(String(st).trim());
+        }
+    }
+    if (Array.isArray(row.case_studies)) {
+        for (const cs of row.case_studies) {
+            const csText = typeof cs === 'string' ? cs : [cs.title || cs.name, cs.snippet || cs.summary].filter(Boolean).join(': ');
+            if (csText)
+                caseStudies.push(csText);
+        }
+    }
+    return {
+        stats: Array.from(new Set(stats)).filter(Boolean),
+        snippets: Array.from(new Set(snippets)).filter(Boolean),
+        caseStudies: Array.from(new Set(caseStudies)).filter(Boolean),
+    };
+}
+// ---------------------------------------------------------------------------
+// Helpers: Graph & Expert Resolution
+// ---------------------------------------------------------------------------
+function normalizeText(text) {
+    return (text || '').toLowerCase().replace(/[^a-z0-9\s]/g, ' ').trim();
+}
+/**
+ * Resolve primary report graph from query string and result rows.
+ */
+export function resolvePrimaryReportGraph(query, resultRows = []) {
+    const allGraphs = getLiveGraphs().length > 0 ? getLiveGraphs() : getGraphs();
+    const reportGraphs = allGraphs.filter(g => g.graph_type === 'industry report' || g.graph_type === 'report');
+    if (reportGraphs.length === 0)
+        return undefined;
+    const normQuery = normalizeText(query);
+    const queryTokens = normQuery.split(/\s+/).filter(t => t.length > 2);
+    // 1. Direct match on organization/curator or graph name in query
+    let bestGraph;
+    let maxMatchScore = 0;
+    for (const g of reportGraphs) {
+        let matchScore = 0;
+        const normName = normalizeText(g.name);
+        const normCurator = normalizeText(g.curator);
+        const normCompany = normalizeText(g.company);
+        const normId = normalizeText(g.graph_id.replace(/-/g, ' '));
+        if (normCurator && normQuery.includes(normCurator))
+            matchScore += 10;
+        if (normCompany && normQuery.includes(normCompany))
+            matchScore += 10;
+        if (normName && normQuery.includes(normName))
+            matchScore += 8;
+        if (normId && normQuery.includes(normId))
+            matchScore += 5;
+        // Token matches
+        for (const token of queryTokens) {
+            if (normName.includes(token))
+                matchScore += 1;
+            if (normCurator.includes(token))
+                matchScore += 2;
+            if (normCompany.includes(token))
+                matchScore += 2;
+        }
+        if (matchScore > maxMatchScore) {
+            maxMatchScore = matchScore;
+            bestGraph = g;
+        }
+    }
+    if (bestGraph && maxMatchScore >= 5) {
+        return bestGraph;
+    }
+    // 2. Resolve from highest density / highest scoring graph in result rows
+    if (resultRows.length > 0) {
+        const graphCounts = new Map();
+        for (const row of resultRows) {
+            const gid = row.graph_id || row.graphId || '';
+            if (!gid)
+                continue;
+            const existing = graphCounts.get(gid) || { count: 0, totalScore: 0 };
+            existing.count += 1;
+            const rawScore = Number(row.relevance_score ?? row.score ?? 0.5);
+            existing.totalScore += rawScore;
+            graphCounts.set(gid, existing);
+        }
+        let topGid = '';
+        let topWeightedScore = -1;
+        for (const [gid, stat] of graphCounts.entries()) {
+            const weighted = stat.count * 2 + stat.totalScore;
+            if (weighted > topWeightedScore) {
+                topWeightedScore = weighted;
+                topGid = gid;
+            }
+        }
+        if (topGid) {
+            const matched = reportGraphs.find(g => g.graph_id === topGid) || getGraphs().find(g => g.graph_id === topGid);
+            if (matched)
+                return matched;
+        }
+    }
+    return bestGraph;
+}
+/**
+ * Filter and match human expert digital twin.
+ * Returns undefined if no genuine topical overlap exists (no arbitrary fallback).
+ */
+export function matchHumanExpertTwin(query, primaryGraph) {
+    const allAnalysts = getAnalysts();
+    const humanTwins = allAnalysts.filter(a => {
+        return a.is_human_agent === true ||
+            a.is_digital_twin === true ||
+            a.isVerifiedRealPerson === true ||
+            a.kind === 'human_agent' ||
+            a.kind === 'human_twin' ||
+            a.type === 'human_agent' ||
+            a.type === 'human_twin' ||
+            a.agent_type === 'human_agent' ||
+            a.agent_type === 'human_twin';
+    });
+    if (humanTwins.length === 0)
+        return undefined;
+    const normQuery = normalizeText(`${query} ${primaryGraph?.name || ''} ${primaryGraph?.domain || ''} ${Array.isArray(primaryGraph?.topics) ? primaryGraph.topics.join(' ') : ''}`);
+    const tokens = normQuery.split(/\s+/).filter(t => t.length > 3);
+    let bestExpert;
+    let maxOverlap = 0;
+    for (const expert of humanTwins) {
+        let score = 0;
+        const expertContext = normalizeText(`${expert.name} ${expert.expert_in || ''} ${expert.what_they_offer || ''} ${Array.isArray(expert.topics) ? expert.topics.join(' ') : ''}`);
+        for (const t of tokens) {
+            if (expertContext.includes(t))
+                score += 1;
+        }
+        if (score > maxOverlap) {
+            maxOverlap = score;
+            bestExpert = expert;
+        }
+    }
+    // Strict: only match when there is positive evidence of topical overlap
+    return maxOverlap > 0 ? bestExpert : undefined;
+}
+/**
+ * Partition result rows into primary report rows vs. other report graph candidate rows.
+ */
+export function partitionReportResults(resultRows, primaryGraphId) {
+    const primaryRows = [];
+    const networkRows = [];
+    const seenPrimaryText = new Set();
+    for (const row of resultRows) {
+        const gid = row.graph_id || row.graphId;
+        const rowText = `${row.name || row.trend_name || ''} ${row.summary || row.description || ''}`;
+        if (primaryGraphId && gid === primaryGraphId) {
+            primaryRows.push(row);
+            seenPrimaryText.add(normalizeText(rowText));
+            const ev = extractEvidenceFromRow(row);
+            for (const s of ev.snippets)
+                seenPrimaryText.add(normalizeText(s));
+            for (const cs of ev.caseStudies)
+                seenPrimaryText.add(normalizeText(cs));
+        }
+        else {
+            networkRows.push(row);
+        }
+    }
+    // If no primary rows were separated (e.g. query spanned multiple graphs equally), split top rows
+    if (primaryRows.length === 0 && resultRows.length > 0) {
+        const topGid = resultRows[0]?.graph_id || resultRows[0]?.graphId;
+        for (const row of resultRows) {
+            const gid = row.graph_id || row.graphId;
+            if (gid === topGid) {
+                primaryRows.push(row);
+                seenPrimaryText.add(normalizeText(`${row.name || row.trend_name || ''} ${row.summary || row.description || ''}`));
+                const ev = extractEvidenceFromRow(row);
+                for (const s of ev.snippets)
+                    seenPrimaryText.add(normalizeText(s));
+                for (const cs of ev.caseStudies)
+                    seenPrimaryText.add(normalizeText(cs));
+            }
+            else {
+                networkRows.push(row);
+            }
+        }
+    }
+    // Server-side deduplication of network rows against primary rows
+    const dedupedNetworkRows = networkRows.filter(nr => {
+        const nrTitle = normalizeText(nr.name || nr.trend_name || '');
+        const nrText = normalizeText(`${nr.name || nr.trend_name || ''} ${nr.summary || nr.description || ''}`);
+        const nrEv = extractEvidenceFromRow(nr);
+        const nrAllTexts = [nrTitle, nrText, ...nrEv.snippets, ...nrEv.caseStudies].map(normalizeText);
+        for (const pText of seenPrimaryText) {
+            for (const text of nrAllTexts) {
+                if (text.length > 15 && (pText.includes(text) || (text.length > 25 && pText.includes(text.slice(0, 25))))) {
+                    return false;
+                }
+            }
+        }
+        return true;
+    });
+    return {
+        primaryRows,
+        networkRows: dedupedNetworkRows,
+    };
+}
+async function runGeminiSynthesis(params) {
+    const primarySummary = params.primaryRows.slice(0, 5).map((r, i) => {
+        const name = r.name || r.trend_name || `Shift ${i + 1}`;
+        const desc = r.summary || r.description || '';
+        const ev = extractEvidenceFromRow(r);
+        const statsStr = ev.stats.slice(0, 3).join(', ');
+        const snipsStr = ev.snippets.slice(0, 2).join('; ');
+        return `[Shift ${i + 1}] "${name}": ${desc}\n  - Stats: ${statsStr || 'None'}\n  - Evidence: ${snipsStr || 'None'}`;
+    }).join('\n');
+    const networkSummary = params.networkRows.slice(0, 4).map((r, i) => {
+        const gid = r.graph_id || r.graphId || `graph-${i}`;
+        const matchedGraph = getLiveGraphs().find(g => g.graph_id === gid) || getGraphs().find(g => g.graph_id === gid);
+        const gName = matchedGraph?.name || r.graph_name || r.graphName || gid;
+        const curator = matchedGraph?.curator || matchedGraph?.company || r.curator || r.organization || 'Research Partner';
+        const name = r.name || r.trend_name || 'Signal';
+        const desc = r.summary || r.description || '';
+        const ev = extractEvidenceFromRow(r);
+        const evSnippet = ev.snippets[0] || desc;
+        return `[Signal ${i + 1}] GraphID: "${gid}" | Report: "${gName}" by ${curator} | Finding: "${name}" - ${evSnippet}`;
+    }).join('\n');
+    const expertContext = params.expert
+        ? `Matched Expert Twin: ${params.expert.name} (Focus: ${params.expert.expert_in || params.expert.what_they_offer || 'Strategy'})`
+        : 'None';
+    const prompt = `---
+title: Fodda 5-Pillar Report Synthesis
+compliance: RFC-2119
+---
+
+### MISSION
+You are an elite research director synthesizing an executive analyst briefing from Fodda's knowledge graph network.
+
+### INPUT DATA
+- User Query: "${params.query}"
+- Primary Report: "${params.reportName}"
+- Primary Report Findings:
+${primarySummary || '(Thin primary findings)'}
+
+- Related Network Signals from other industry reports:
+${networkSummary || '(No other report signals)'}
+
+- ${expertContext}
+
+### REQUIREMENTS
+Produce a single JSON object with these exact keys:
+1. "topline_hook": A 1-2 sentence provocative thesis and core tension stopping the reader. Do NOT lead with dry bibliographic metadata.
+2. "shifts": Array of 3-5 structural shifts. Each must have:
+   - "title": Sharp, evocative shift title.
+   - "narrative": 1-2 sentences explaining why this shift matters.
+   - "stats": Array of 1-3 concrete data points, metrics, or percentages extracted from the evidence.
+   - "evidence": Array of 1-2 concrete proof points or case study mentions.
+3. "network_signals": Array of 2-3 cross-graph connections. For each signal:
+   - "graph_id": The exact GraphID provided in the input.
+   - "signal_type": MUST be either "validates", "contrasts", or "related".
+   - "connection": 1 crisp sentence explaining how this other report supports or contrasts the primary finding.
+4. "expert_stance": ${params.expert ? 'A 1-sentence analytical perspective in the voice/philosophy of the matched expert twin regarding this topic.' : 'null'}
+
+STRICT RULE: Do NOT mention tokens, shared payment tokens, SPT, or pricing mechanics.`;
+    try {
+        const resp = await callGemini(prompt, 'gemini-2.0-flash', 1500);
+        if (!resp.text || resp.error)
+            return null;
+        let parsed;
+        try {
+            parsed = JSON.parse(resp.text);
+        }
+        catch {
+            return null;
+        }
+        if (typeof parsed === 'object' && parsed !== null) {
+            return parsed;
+        }
+        return null;
+    }
+    catch {
+        return null;
+    }
+}
+// ---------------------------------------------------------------------------
+// Pre-Rendered Markdown Formatter
+// ---------------------------------------------------------------------------
+function renderBriefingMarkdown(payload) {
+    const lines = [];
+    lines.push(`# Strategic Briefing: ${payload.reportTitle}`);
+    lines.push('');
+    lines.push(`> **Core Tension:** ${payload.toplineHook}`);
+    lines.push('');
+    lines.push('---');
+    lines.push('');
+    // Pillar 2: Core Shifts
+    lines.push('### 📊 Core Structural Shifts & Evidence');
+    lines.push('');
+    for (const shift of payload.shifts) {
+        lines.push(`#### • ${shift.title}`);
+        lines.push(`${shift.narrative}`);
+        if (shift.stats.length > 0) {
+            lines.push('');
+            for (const stat of shift.stats) {
+                lines.push(`  - **Key Metric:** ${stat}`);
+            }
+        }
+        if (shift.evidence.length > 0) {
+            lines.push('');
+            for (const ev of shift.evidence) {
+                lines.push(`  - **Evidence:** ${ev}`);
+            }
+        }
+        lines.push('');
+    }
+    // Pillar 3: Cross-Graph Network Evidence
+    if (payload.networkSignals.length > 0) {
+        lines.push('---');
+        lines.push('');
+        lines.push('### 🌐 Cross-Graph Network Signals');
+        lines.push('');
+        for (const sig of payload.networkSignals) {
+            const badge = sig.signal_type === 'validates'
+                ? '✅ **Validates**'
+                : (sig.signal_type === 'contrasts' ? '⚡ **Contrasts**' : '🔗 **Related Signal**');
+            const curatorInfo = sig.curator ? ` (${sig.curator})` : '';
+            lines.push(`- ${badge} → [${sig.source_report}](${sig.url})${curatorInfo}: ${sig.snippet}`);
+        }
+        lines.push('');
+    }
+    // Pillar 4: Expert Spotlight
+    if (payload.expertSpotlight) {
+        lines.push('---');
+        lines.push('');
+        lines.push('### 👤 Human Expert Digital Twin Spotlight');
+        lines.push('');
+        lines.push(`**Expert Alignment:** ${payload.expertSpotlight.name}`);
+        lines.push(`*Why Matched:* ${payload.expertSpotlight.why_matched}`);
+        lines.push(`> *"${payload.expertSpotlight.position}"*`);
+        if (payload.expertSpotlight.consult_url) {
+            lines.push(`→ **Consult ${payload.expertSpotlight.name}'s Twin:** [${payload.expertSpotlight.consult_url}](${payload.expertSpotlight.consult_url})`);
+        }
+        lines.push('');
+    }
+    // Pillar 5: Follow-Up Prompts
+    if (payload.followUps.length > 0) {
+        lines.push('---');
+        lines.push('');
+        lines.push('### 🎯 Suggested Deep-Dive Moves');
+        lines.push('');
+        for (const fu of payload.followUps) {
+            lines.push(`- **${fu.prompt}**  `);
+            lines.push(`  *Run Tool:* \`${fu.tool}(${JSON.stringify(fu.args)})\` | [Open in Fodda](${fu.app_url})`);
+        }
+        lines.push('');
+    }
+    return lines.join('\n').trim();
+}
+// ---------------------------------------------------------------------------
+// Main Builder: buildReportEditorialBriefing
+// ---------------------------------------------------------------------------
+export async function buildReportEditorialBriefing(params) {
+    const { data, query, annotatedData } = params;
+    const rawRows = Array.isArray(data?.results)
+        ? data.results
+        : (Array.isArray(data?.trends) ? data.trends : (Array.isArray(data?.data) ? data.data : []));
+    // 1. Resolve Primary Graph & Partition Rows
+    const primaryGraph = resolvePrimaryReportGraph(query, rawRows);
+    const primaryGraphId = primaryGraph?.graph_id;
+    const reportTitle = primaryGraph?.name || (rawRows[0]?.graph_name) || (rawRows[0]?.graph_id) || 'Industry Report Intelligence';
+    const { primaryRows, networkRows } = partitionReportResults(rawRows, primaryGraphId);
+    // 2. Pre-computed Ingestion-Time 5-Pillar Metadata (from CE / API)
+    const graphEditorial = (primaryGraphId && data?.graph_editorial?.[primaryGraphId]) || {};
+    const precomputedCoreTension = primaryGraph?.core_tension || graphEditorial.core_tension || rawRows[0]?.core_tension;
+    const precomputedSignals = primaryGraph?.cross_graph_signals || graphEditorial.cross_graph_signals || rawRows[0]?.cross_graph_signals;
+    const precomputedTwinSlug = primaryGraph?.matched_human_twin_slug || graphEditorial.matched_human_twin_slug || rawRows[0]?.matched_human_twin_slug;
+    const precomputedExpertQuote = primaryGraph?.expert_stance_quote || graphEditorial.expert_stance_quote || rawRows[0]?.expert_stance_quote;
+    // 3. Match Human Expert Digital Twin (Prefer pre-computed twin slug if available)
+    let matchedExpert;
+    if (precomputedTwinSlug) {
+        matchedExpert = getAnalysts().find(a => (a.slug || a.expertSlug || a.analyst_id) === precomputedTwinSlug);
+    }
+    if (!matchedExpert) {
+        matchedExpert = matchHumanExpertTwin(query, primaryGraph);
+    }
+    // 4. Single Unified Gemini Synthesis Call (if any fields need LLM fill)
+    const needsGemini = !precomputedCoreTension || !precomputedSignals || precomputedSignals.length === 0;
+    const geminiOutput = needsGemini ? await runGeminiSynthesis({
+        query,
+        reportName: reportTitle,
+        primaryRows,
+        networkRows,
+        expert: matchedExpert,
+    }) : null;
+    // 5. Construct Topline Hook (Pillar 1)
+    let toplineHookText = precomputedCoreTension || geminiOutput?.topline_hook?.slice(0, 350);
+    let isHookGenerated = !precomputedCoreTension && Boolean(geminiOutput?.topline_hook);
+    if (!toplineHookText) {
+        isHookGenerated = false;
+        if (primaryRows.length > 0) {
+            const topRow = primaryRows[0];
+            const name = topRow?.name || topRow?.trend_name || 'Emerging Dynamics';
+            const sum = topRow?.summary || topRow?.description || 'key strategic shifts across the sector';
+            toplineHookText = `Findings from ${reportTitle} indicate that ${name} is driving ${sum}.`;
+        }
+        else {
+            toplineHookText = `No published report intelligence was found specifically matching "${query}". Exploring adjacent domain signals or consulting a specialist analyst can help bridge this gap.`;
+        }
+    }
+    // 5. Construct Structural Shifts (Pillar 2)
+    const shifts = [];
+    if (geminiOutput?.shifts && Array.isArray(geminiOutput.shifts) && geminiOutput.shifts.length > 0) {
+        for (let i = 0; i < Math.min(geminiOutput.shifts.length, 4); i++) {
+            const s = geminiOutput.shifts[i];
+            if (!s)
+                continue;
+            const matchingRow = primaryRows[i] || primaryRows[0];
+            const rowEv = matchingRow ? extractEvidenceFromRow(matchingRow) : { stats: [], snippets: [], caseStudies: [] };
+            const finalStats = Array.isArray(s.stats) && s.stats.length > 0
+                ? s.stats.slice(0, 3)
+                : rowEv.stats.slice(0, 3);
+            const finalEvidence = Array.isArray(s.evidence) && s.evidence.length > 0
+                ? s.evidence.slice(0, 2)
+                : rowEv.snippets.slice(0, 2);
+            shifts.push({
+                title: (s.title || `Shift ${i + 1}`).slice(0, 100),
+                narrative: (s.narrative || '').slice(0, 300),
+                stats: finalStats,
+                evidence: finalEvidence,
+                generated: true,
+            });
+        }
+    }
+    else {
+        // Fallback: Mechanical extraction from primary rows
+        const fallbackRows = (primaryRows.length > 0 ? primaryRows : rawRows).slice(0, 4);
+        for (let i = 0; i < fallbackRows.length; i++) {
+            const row = fallbackRows[i];
+            const title = (row?.name || row?.trend_name || `Shift ${i + 1}`).slice(0, 100);
+            const narrative = (row?.summary || row?.description || 'Structural shift identified in report data.').slice(0, 300);
+            const ev = extractEvidenceFromRow(row);
+            shifts.push({
+                title,
+                narrative,
+                stats: ev.stats.slice(0, 3),
+                evidence: ev.snippets.slice(0, 2),
+                generated: false,
+            });
+        }
+    }
+    // 6. Construct Network Signals (Pillar 3)
+    const networkSignals = [];
+    if (precomputedSignals && Array.isArray(precomputedSignals) && precomputedSignals.length > 0) {
+        for (const sig of precomputedSignals.slice(0, 3)) {
+            const gid = sig.graph_id || '';
+            const matchedGraph = getLiveGraphs().find(g => g.graph_id === gid) || getGraphs().find(g => g.graph_id === gid);
+            const sourceReport = matchedGraph?.name || gid || 'Industry Report';
+            const curator = matchedGraph?.curator || matchedGraph?.company || 'Research Partner';
+            networkSignals.push({
+                title: sig.source_case_study || `Signal from ${sourceReport}`,
+                source_report: sourceReport,
+                graph_id: gid,
+                curator: curator ? String(curator) : undefined,
+                signal_type: sig.signal_type || 'validates',
+                snippet: (sig.connection_summary || sig.source_case_study || '').slice(0, 250),
+                url: matchedGraph?.webpage_url || `https://app.fodda.ai/graphs/${encodeURIComponent(gid)}`,
+            });
+        }
+    }
+    else {
+        const candidateNetRows = networkRows.slice(0, 3);
+        for (let i = 0; i < candidateNetRows.length; i++) {
+            const nr = candidateNetRows[i];
+            const gid = nr?.graph_id || nr?.graphId || '';
+            const matchedGraph = getLiveGraphs().find(g => g.graph_id === gid) || getGraphs().find(g => g.graph_id === gid);
+            const sourceReport = matchedGraph?.name || nr?.graph_name || nr?.graphName || gid || 'Industry Report';
+            const curator = matchedGraph?.curator || matchedGraph?.company || nr?.curator || nr?.organization || 'Research Partner';
+            // Match LLM signal classification if available
+            const matchedGeminiSig = geminiOutput?.network_signals?.find(s => s.graph_id === gid || s.graph_id === sourceReport);
+            const signalType = matchedGeminiSig?.signal_type || 'related';
+            const ev = extractEvidenceFromRow(nr);
+            const snippet = (matchedGeminiSig?.connection || ev.snippets[0] || nr?.summary || nr?.description || nr?.name || nr?.trend_name || '').slice(0, 250);
+            const rawScore = Number(nr?.relevance_score ?? nr?.score);
+            networkSignals.push({
+                title: nr?.name || nr?.trend_name || `Signal from ${sourceReport}`,
+                source_report: sourceReport,
+                graph_id: gid,
+                curator: curator ? String(curator) : undefined,
+                signal_type: signalType,
+                snippet,
+                score: !isNaN(rawScore) ? rawScore : undefined,
+                url: matchedGraph?.webpage_url || `https://app.fodda.ai/graphs/${encodeURIComponent(gid)}`,
+            });
+        }
+    }
+    // 7. Construct Expert Spotlight (Pillar 4)
+    let expertSpotlight;
+    if (matchedExpert) {
+        const rawSlug = matchedExpert.slug || matchedExpert.expertSlug || matchedExpert.analyst_id || '';
+        const isCleanSlug = /^[a-z0-9-]+$/.test(rawSlug);
+        const consultUrl = isCleanSlug ? `https://expert.fodda.ai/${rawSlug}` : undefined;
+        const position = (precomputedExpertQuote || geminiOutput?.expert_stance || matchedExpert.what_they_offer || `Offers strategic advisory on ${matchedExpert.expert_in || 'market shifts'}.`).slice(0, 250);
+        const whyMatched = `Specializes in ${matchedExpert.expert_in || matchedExpert.name}'s strategic domain`;
+        expertSpotlight = {
+            name: matchedExpert.name,
+            slug: isCleanSlug ? rawSlug : undefined,
+            analyst_id: matchedExpert.analyst_id,
+            why_matched: whyMatched,
+            position,
+            consult_url: consultUrl,
+            consult_tool: 'consult_human_agent',
+            generated: !precomputedExpertQuote && Boolean(geminiOutput?.expert_stance),
+        };
+    }
+    // 8. Reshape Follow-Ups (Pillar 5)
+    const followUps = [];
+    // Follow-up 1: Deep dive on top shift or primary report
+    if (shifts.length > 0 && shifts[0]) {
+        const topShift = shifts[0];
+        followUps.push({
+            prompt: `Deep dive into "${topShift.title}" in ${reportTitle}`,
+            tool: 'get_report_intelligence',
+            args: { query: `${reportTitle} ${topShift.title}` },
+            app_url: primaryGraphId
+                ? `https://app.fodda.ai/graphs/${encodeURIComponent(primaryGraphId)}`
+                : `https://app.fodda.ai`,
+        });
+    }
+    // Follow-up 2: Consult matched human twin
+    if (matchedExpert) {
+        followUps.push({
+            prompt: `Consult ${matchedExpert.name} on strategic execution`,
+            tool: 'consult_human_agent',
+            args: {
+                analyst_id: matchedExpert.analyst_id,
+                query: `How does ${reportTitle} impact strategic priorities for brands?`,
+            },
+            app_url: expertSpotlight?.consult_url || `https://app.fodda.ai`,
+        });
+    }
+    // Follow-up 3: Cross-graph query
+    if (networkSignals.length > 0 && networkSignals[0]) {
+        const netSig = networkSignals[0];
+        followUps.push({
+            prompt: `Explore cross-industry validation from ${netSig.source_report}`,
+            tool: 'get_report_intelligence',
+            args: { query: `${netSig.source_report} ${query}` },
+            app_url: netSig.url,
+        });
+    }
+    // If fewer than 3 follow-ups generated, pull from annotatedData.next_moves candidates if available
+    if (followUps.length < 3 && annotatedData?.next_moves?.candidates && Array.isArray(annotatedData.next_moves.candidates)) {
+        for (const c of annotatedData.next_moves.candidates) {
+            if (followUps.length >= 3)
+                break;
+            if (c.tool && c.prompt) {
+                followUps.push({
+                    prompt: c.prompt,
+                    tool: c.tool,
+                    args: c.args || {},
+                    app_url: 'https://app.fodda.ai',
+                });
+            }
+        }
+    }
+    // 9. Pre-Render Markdown
+    const briefingMarkdown = renderBriefingMarkdown({
+        reportTitle,
+        toplineHook: toplineHookText,
+        shifts,
+        networkSignals,
+        expertSpotlight,
+        followUps,
+    });
+    return {
+        view: 'editorial',
+        query,
+        report_title: reportTitle,
+        primary_graph_id: primaryGraphId,
+        topline_hook: {
+            text: toplineHookText,
+            generated: isHookGenerated,
+        },
+        shifts,
+        network_signals: networkSignals,
+        expert_spotlight: expertSpotlight,
+        follow_ups: followUps,
+        briefing_markdown: briefingMarkdown,
+        coverage: annotatedData?.coverage || data?.coverage,
+    };
+}
+//# sourceMappingURL=reportBriefing.js.map