@mathrunet/masamune_firebase_workflow_marketing 3.8.0

package/dist/functions/analyze_marketing_data.js

@@ -0,0 +1,990 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.AnalyzeMarketingData = void 0;
+const masamune_firebase_workflow_1 = require("@mathrunet/masamune_firebase_workflow");
+const genai_1 = require("@google/genai");
+const locales_1 = require("../locales");
+/**
+ * A function for analyzing marketing data using AI (Gemini).
+ *
+ * AIを使用してマーケティングデータを解析するためのFunction。
+ */
+class AnalyzeMarketingData extends masamune_firebase_workflow_1.WorkflowProcessFunctionBase {
+    /**
+     * The ID of the function.
+     *
+     * 関数のID。
+     */
+    id = "analyze_marketing_data";
+    /**
+     * Translations for the current locale.
+     */
+    translations = (0, locales_1.getTranslations)("en");
+    /**
+     * The process of the function.
+     *
+     * @param context
+     * The context of the function.
+     *
+     * @returns
+     * The action of the function.
+     */
+    async process(context) {
+        const action = context.action;
+        const task = context.task;
+        // 1. task.results から各データを取得
+        const googlePlayConsole = task.results?.googlePlayConsole;
+        const appStore = task.results?.appStore;
+        const firebaseAnalytics = task.results?.firebaseAnalytics;
+        const githubRepository = task.results?.githubRepository;
+        // 市場調査データを取得（あれば）
+        const marketResearchData = task.results?.marketResearchData;
+        const marketResearch = task.results?.marketResearch;
+        // 市場調査データが有効かどうかをチェック
+        const hasValidMarketResearchData = marketResearchData && !("error" in marketResearchData);
+        const hasValidMarketResearch = marketResearch && !("error" in marketResearch);
+        // 2. いずれのデータも無ければ空データを返却
+        if (!googlePlayConsole && !appStore && !firebaseAnalytics) {
+            console.log("AnalyzeMarketingData: No marketing data found in task.results");
+            return {
+                ...action,
+                results: {
+                    marketingAnalytics: {},
+                }
+            };
+        }
+        // 3. 環境変数からプロジェクト情報を取得
+        const projectId = process.env.GCLOUD_PROJECT || process.env.GOOGLE_CLOUD_PROJECT || process.env.GCP_PROJECT_ID;
+        const region = process.env.GCLOUD_REGION || "us-central1";
+        const modelName = process.env.GEMINI_MODEL || "gemini-2.5-flash";
+        const inputPrice = process.env.MODEL_INPUT_PRICE || 0.0000003;
+        const outputPrice = process.env.MODEL_OUTPUT_PRICE || 0.0000025;
+        if (!projectId) {
+            console.error("AnalyzeMarketingData: No GCP project ID found");
+            return {
+                ...action,
+                results: {
+                    marketingAnalytics: {
+                        error: "No GCP project ID configured",
+                    },
+                }
+            };
+        }
+        try {
+            // 4. Localeを取得して翻訳を初期化
+            const locale = typeof action.locale === "object"
+                ? action.locale["@language"]
+                : action.locale;
+            this.translations = (0, locales_1.getTranslations)(locale);
+            // 5. GoogleGenAI を初期化（VertexAI モード）
+            const genai = new genai_1.GoogleGenAI({
+                vertexai: true,
+                project: projectId,
+                location: region,
+            });
+            const combinedData = {
+                googlePlayConsole,
+                appStore,
+                firebaseAnalytics,
+                githubRepository,
+                marketResearchData: hasValidMarketResearchData ? marketResearchData : undefined,
+                marketResearch: hasValidMarketResearch ? marketResearch : undefined,
+            };
+            // 市場調査データが統合されるかどうかのフラグ（Firestoreはundefinedを許可しないため、必ずbooleanにする）
+            const marketDataIntegrated = !!(hasValidMarketResearchData || hasValidMarketResearch);
+            // レビューデータを抽出
+            const reviews = [
+                ...(googlePlayConsole?.recentReviews || []),
+                ...(appStore?.recentReviews || []),
+            ];
+            // 価格を数値に変換（環境変数は文字列の可能性があるため）
+            const inputPriceNum = Number(inputPrice);
+            const outputPriceNum = Number(outputPrice);
+            // 5. 各解析を並列実行
+            const [overallResult, suggestionsResult, trendResult, reviewResult, githubImprovementsResult, competitivePositioningResult, marketOpportunityResult,] = await Promise.all([
+                this.generateOverallAnalysis(genai, combinedData, modelName),
+                this.generateImprovementSuggestions(genai, combinedData, modelName),
+                this.generateTrendAnalysis(genai, combinedData, modelName),
+                reviews.length > 0
+                    ? this.analyzeReviews(genai, reviews, modelName)
+                    : Promise.resolve({
+                        data: {
+                            sentiment: { positive: 0, neutral: 0, negative: 0 },
+                            commonThemes: [],
+                            actionableInsights: [],
+                        },
+                        inputTokens: 0,
+                        outputTokens: 0,
+                    }),
+                // GitHub改善提案（GitHubデータがある場合のみ）
+                githubRepository && !("error" in githubRepository)
+                    ? this.generateGitHubImprovements(genai, combinedData, modelName)
+                    : Promise.resolve(null),
+                // 競合ポジショニング分析（市場調査データがある場合のみ）
+                marketDataIntegrated
+                    ? this.generateCompetitivePositioning(genai, combinedData, modelName)
+                    : Promise.resolve(null),
+                // 市場機会優先度分析（市場調査データがある場合のみ）
+                marketDataIntegrated
+                    ? this.generateMarketOpportunityPriority(genai, combinedData, modelName)
+                    : Promise.resolve(null),
+            ]);
+            // 6. トークン使用量を集計してコストを計算
+            const totalInputTokens = overallResult.inputTokens +
+                suggestionsResult.inputTokens +
+                trendResult.inputTokens +
+                reviewResult.inputTokens +
+                (githubImprovementsResult?.inputTokens ?? 0) +
+                (competitivePositioningResult?.inputTokens ?? 0) +
+                (marketOpportunityResult?.inputTokens ?? 0);
+            const totalOutputTokens = overallResult.outputTokens +
+                suggestionsResult.outputTokens +
+                trendResult.outputTokens +
+                reviewResult.outputTokens +
+                (githubImprovementsResult?.outputTokens ?? 0) +
+                (competitivePositioningResult?.outputTokens ?? 0) +
+                (marketOpportunityResult?.outputTokens ?? 0);
+            // コスト計算（1トークンあたりのドル × トークン数）
+            const aiCost = totalInputTokens * inputPriceNum + totalOutputTokens * outputPriceNum;
+            // 7. 結果を marketingAnalytics キーで返却
+            return {
+                ...action,
+                usage: (action.usage ?? 0) + aiCost,
+                results: {
+                    marketingAnalytics: {
+                        overallAnalysis: overallResult.data,
+                        improvementSuggestions: suggestionsResult.data,
+                        trendAnalysis: trendResult.data,
+                        reviewAnalysis: reviewResult.data,
+                        // 市場調査データがある場合のみ追加
+                        ...(competitivePositioningResult?.data ? { competitivePositioning: competitivePositioningResult.data } : {}),
+                        ...(marketOpportunityResult?.data ? { marketOpportunityPriority: marketOpportunityResult.data } : {}),
+                        marketDataIntegrated,
+                        generatedAt: new Date().toISOString(),
+                    },
+                    // GitHub改善提案（GitHubデータがある場合のみ）
+                    ...(githubImprovementsResult?.data ? { githubImprovements: githubImprovementsResult.data } : {}),
+                }
+            };
+        }
+        catch (error) {
+            console.error("AnalyzeMarketingData: Failed to analyze data", error);
+            return {
+                ...action,
+                results: {
+                    marketingAnalytics: {
+                        error: error.message,
+                    },
+                }
+            };
+        }
+    }
+    /**
+     * Generate overall analysis from marketing data.
+     */
+    async generateOverallAnalysis(genai, data, modelName) {
+        try {
+            const prompt = this.buildOverallAnalysisPrompt(data);
+            const response = await genai.models.generateContent({
+                model: modelName,
+                contents: prompt,
+                config: {
+                    responseMimeType: "application/json",
+                    responseSchema: {
+                        type: genai_1.Type.OBJECT,
+                        properties: {
+                            summary: { type: genai_1.Type.STRING },
+                            highlights: {
+                                type: genai_1.Type.ARRAY,
+                                items: { type: genai_1.Type.STRING },
+                            },
+                            concerns: {
+                                type: genai_1.Type.ARRAY,
+                                items: { type: genai_1.Type.STRING },
+                            },
+                            keyMetrics: {
+                                type: genai_1.Type.ARRAY,
+                                items: {
+                                    type: genai_1.Type.OBJECT,
+                                    properties: {
+                                        metric: { type: genai_1.Type.STRING },
+                                        value: { type: genai_1.Type.STRING },
+                                        trend: { type: genai_1.Type.STRING },
+                                    },
+                                },
+                            },
+                        },
+                        required: ["summary", "highlights", "concerns", "keyMetrics"],
+                    },
+                },
+            });
+            const text = response.text;
+            if (!text) {
+                throw new Error("No response from AI model");
+            }
+            const inputTokens = response.usageMetadata?.promptTokenCount ?? 0;
+            const outputTokens = response.usageMetadata?.candidatesTokenCount ?? 0;
+            return {
+                data: JSON.parse(text),
+                inputTokens,
+                outputTokens,
+            };
+        }
+        catch (err) {
+            console.error("Failed to generate overall analysis:", err.message);
+            return {
+                data: {
+                    summary: "",
+                    highlights: [],
+                    concerns: [],
+                    keyMetrics: [],
+                },
+                inputTokens: 0,
+                outputTokens: 0,
+            };
+        }
+    }
+    /**
+     * Generate improvement suggestions.
+     */
+    async generateImprovementSuggestions(genai, data, modelName) {
+        try {
+            const prompt = this.buildImprovementSuggestionsPrompt(data);
+            const response = await genai.models.generateContent({
+                model: modelName,
+                contents: prompt,
+                config: {
+                    responseMimeType: "application/json",
+                    responseSchema: {
+                        type: genai_1.Type.ARRAY,
+                        items: {
+                            type: genai_1.Type.OBJECT,
+                            properties: {
+                                title: { type: genai_1.Type.STRING },
+                                description: { type: genai_1.Type.STRING },
+                                priority: { type: genai_1.Type.STRING },
+                                category: { type: genai_1.Type.STRING },
+                                expectedImpact: { type: genai_1.Type.STRING },
+                            },
+                            required: ["title", "description", "priority", "category"],
+                        },
+                    },
+                },
+            });
+            const text = response.text;
+            if (!text) {
+                throw new Error("No response from AI model");
+            }
+            const inputTokens = response.usageMetadata?.promptTokenCount ?? 0;
+            const outputTokens = response.usageMetadata?.candidatesTokenCount ?? 0;
+            return {
+                data: JSON.parse(text),
+                inputTokens,
+                outputTokens,
+            };
+        }
+        catch (err) {
+            console.error("Failed to generate improvement suggestions:", err.message);
+            return {
+                data: [],
+                inputTokens: 0,
+                outputTokens: 0,
+            };
+        }
+    }
+    /**
+     * Generate trend analysis.
+     */
+    async generateTrendAnalysis(genai, data, modelName) {
+        try {
+            const prompt = this.buildTrendAnalysisPrompt(data);
+            const response = await genai.models.generateContent({
+                model: modelName,
+                contents: prompt,
+                config: {
+                    responseMimeType: "application/json",
+                    responseSchema: {
+                        type: genai_1.Type.OBJECT,
+                        properties: {
+                            userGrowthTrend: { type: genai_1.Type.STRING },
+                            engagementTrend: { type: genai_1.Type.STRING },
+                            ratingTrend: { type: genai_1.Type.STRING },
+                            predictions: {
+                                type: genai_1.Type.ARRAY,
+                                items: { type: genai_1.Type.STRING },
+                            },
+                        },
+                        required: ["userGrowthTrend", "engagementTrend", "ratingTrend", "predictions"],
+                    },
+                },
+            });
+            const text = response.text;
+            if (!text) {
+                throw new Error("No response from AI model");
+            }
+            const inputTokens = response.usageMetadata?.promptTokenCount ?? 0;
+            const outputTokens = response.usageMetadata?.candidatesTokenCount ?? 0;
+            return {
+                data: JSON.parse(text),
+                inputTokens,
+                outputTokens,
+            };
+        }
+        catch (err) {
+            console.error("Failed to generate trend analysis:", err.message);
+            return {
+                data: {
+                    userGrowthTrend: "",
+                    engagementTrend: "",
+                    ratingTrend: "",
+                    predictions: [],
+                },
+                inputTokens: 0,
+                outputTokens: 0,
+            };
+        }
+    }
+    /**
+     * Analyze reviews and extract insights.
+     */
+    async analyzeReviews(genai, reviews, modelName) {
+        try {
+            const prompt = this.buildReviewAnalysisPrompt(reviews);
+            const response = await genai.models.generateContent({
+                model: modelName,
+                contents: prompt,
+                config: {
+                    responseMimeType: "application/json",
+                    responseSchema: {
+                        type: genai_1.Type.OBJECT,
+                        properties: {
+                            sentiment: {
+                                type: genai_1.Type.OBJECT,
+                                properties: {
+                                    positive: { type: genai_1.Type.NUMBER },
+                                    neutral: { type: genai_1.Type.NUMBER },
+                                    negative: { type: genai_1.Type.NUMBER },
+                                },
+                            },
+                            commonThemes: {
+                                type: genai_1.Type.ARRAY,
+                                items: { type: genai_1.Type.STRING },
+                            },
+                            actionableInsights: {
+                                type: genai_1.Type.ARRAY,
+                                items: { type: genai_1.Type.STRING },
+                            },
+                        },
+                        required: ["sentiment", "commonThemes", "actionableInsights"],
+                    },
+                },
+            });
+            const text = response.text;
+            if (!text) {
+                throw new Error("No response from AI model");
+            }
+            const inputTokens = response.usageMetadata?.promptTokenCount ?? 0;
+            const outputTokens = response.usageMetadata?.candidatesTokenCount ?? 0;
+            return {
+                data: JSON.parse(text),
+                inputTokens,
+                outputTokens,
+            };
+        }
+        catch (err) {
+            console.error("Failed to analyze reviews:", err.message);
+            return {
+                data: {
+                    sentiment: { positive: 0, neutral: 0, negative: 0 },
+                    commonThemes: [],
+                    actionableInsights: [],
+                },
+                inputTokens: 0,
+                outputTokens: 0,
+            };
+        }
+    }
+    /**
+     * Build prompt for overall analysis.
+     */
+    buildOverallAnalysisPrompt(data) {
+        // 市場調査データのセクションを構築
+        let marketResearchSection = "";
+        if (data.marketResearchData || data.marketResearch) {
+            marketResearchSection = `
+## 市場調査データ
+${data.marketResearchData ? `
+### 市場ポテンシャル
+${JSON.stringify(data.marketResearchData.marketPotential, null, 2)}
+
+### 競合分析
+${JSON.stringify(data.marketResearchData.competitorAnalysis, null, 2)}
+` : ""}
+${data.marketResearch ? `
+### 需要予測
+${JSON.stringify(data.marketResearch.demandForecast, null, 2)}
+
+### キーインサイト
+${JSON.stringify(data.marketResearch.keyInsights, null, 2)}
+` : ""}
+`;
+        }
+        return `You are an expert app marketing analyst. Analyze the following marketing data and provide a comprehensive overview.
+
+## Marketing Data
+
+### Google Play Data
+${data.googlePlayConsole ? JSON.stringify(data.googlePlayConsole, null, 2) : "Not available"}
+
+### App Store Data
+${data.appStore ? JSON.stringify(data.appStore, null, 2) : "Not available"}
+
+### Firebase Analytics Data
+${data.firebaseAnalytics ? JSON.stringify(data.firebaseAnalytics, null, 2) : "Not available"}
+${marketResearchSection}
+## Instructions
+1. Write a concise summary (2-3 paragraphs) analyzing the overall app performance
+2. List 3-5 key highlights (positive points)
+3. List 2-4 concerns or areas needing attention
+4. Identify 4-6 key metrics with their values and trends (up/down/stable)
+${data.marketResearchData || data.marketResearch ? "5. Integrate market research insights into your analysis, considering market potential and competitive landscape" : ""}
+
+${this.translations.respondInLanguage}`;
+    }
+    /**
+     * Build prompt for improvement suggestions.
+     */
+    buildImprovementSuggestionsPrompt(data) {
+        // 市場調査データのセクションを構築
+        let marketResearchSection = "";
+        if (data.marketResearchData || data.marketResearch) {
+            marketResearchSection = `
+## 市場調査データ
+${data.marketResearchData ? `
+### ビジネス機会
+${JSON.stringify(data.marketResearchData.businessOpportunities, null, 2)}
+
+### 差別化機会
+${JSON.stringify(data.marketResearchData.competitorAnalysis.differentiationOpportunities, null, 2)}
+
+### 市場ギャップ
+${JSON.stringify(data.marketResearchData.competitorAnalysis.marketGaps, null, 2)}
+` : ""}
+${data.marketResearch ? `
+### 収益向上施策（市場調査ベース）
+${JSON.stringify(data.marketResearch.revenueStrategies, null, 2)}
+
+### 流入向上施策（市場調査ベース）
+${JSON.stringify(data.marketResearch.trafficStrategies, null, 2)}
+` : ""}
+`;
+        }
+        return `You are an expert app marketing strategist. Based on the following marketing data, provide actionable improvement suggestions.
+
+## Marketing Data
+
+### Google Play Data
+${data.googlePlayConsole ? JSON.stringify(data.googlePlayConsole, null, 2) : "Not available"}
+
+### App Store Data
+${data.appStore ? JSON.stringify(data.appStore, null, 2) : "Not available"}
+
+### Firebase Analytics Data
+${data.firebaseAnalytics ? JSON.stringify(data.firebaseAnalytics, null, 2) : "Not available"}
+${marketResearchSection}
+## Instructions
+Provide 5-8 specific, actionable improvement suggestions. For each suggestion:
+1. title: Brief title (under 50 characters)
+2. description: Detailed explanation with specific steps (100-200 characters)
+3. priority: "high", "medium", or "low"
+4. category: One of "user_acquisition", "retention", "engagement", "monetization", "quality", "development"
+5. expectedImpact: Expected outcome if implemented
+${data.marketResearchData || data.marketResearch ? `
+Important: Consider the market research data when making recommendations:
+- Prioritize suggestions that address identified market gaps
+- Leverage business opportunities and differentiation points
+- Align with revenue and traffic strategies from market research
+` : ""}
+Focus on data-driven recommendations. ${this.translations.respondInLanguage}`;
+    }
+    /**
+     * Build prompt for trend analysis.
+     */
+    buildTrendAnalysisPrompt(data) {
+        // 市場調査データのセクションを構築
+        let marketResearchSection = "";
+        if (data.marketResearchData || data.marketResearch) {
+            marketResearchSection = `
+## 市場調査データ
+${data.marketResearch ? `
+### 需要予測
+${JSON.stringify(data.marketResearch.demandForecast, null, 2)}
+
+### 市場全体トレンド
+${data.marketResearch.demandForecast.overallTrend}
+` : ""}
+${data.marketResearchData ? `
+### 市場ドライバー
+${JSON.stringify(data.marketResearchData.marketPotential.marketDrivers, null, 2)}
+
+### 市場ポテンシャル概要
+${data.marketResearchData.marketPotential.summary}
+` : ""}
+`;
+        }
+        return `You are an expert data analyst specializing in mobile app trends. Analyze the following marketing data and provide trend insights.
+
+## Marketing Data
+
+### Google Play Data
+${data.googlePlayConsole ? JSON.stringify(data.googlePlayConsole, null, 2) : "Not available"}
+
+### App Store Data
+${data.appStore ? JSON.stringify(data.appStore, null, 2) : "Not available"}
+
+### Firebase Analytics Data
+${data.firebaseAnalytics ? JSON.stringify(data.firebaseAnalytics, null, 2) : "Not available"}
+${marketResearchSection}
+## Instructions
+Analyze trends and provide:
+1. userGrowthTrend: Analysis of user acquisition and growth patterns (2-3 sentences)
+2. engagementTrend: Analysis of user engagement metrics (2-3 sentences)
+3. ratingTrend: Analysis of app ratings and user satisfaction (2-3 sentences)
+4. predictions: 3-5 predictions for the next period based on current trends
+${data.marketResearchData || data.marketResearch ? "5. Consider market demand forecasts and market drivers when making predictions" : ""}
+
+${this.translations.respondInLanguage}`;
+    }
+    /**
+     * Build prompt for review analysis.
+     */
+    buildReviewAnalysisPrompt(reviews) {
+        const reviewsText = reviews
+            .slice(0, 50) // Limit to 50 reviews to avoid token limits
+            .map((r) => `Rating: ${r.rating}/5\nText: ${r.text}\n---`)
+            .join("\n");
+        return `You are an expert in user feedback analysis. Analyze the following app reviews and extract insights.
+
+## Reviews
+${reviewsText}
+
+## Instructions
+Analyze the reviews and provide:
+1. sentiment: Percentage breakdown of positive, neutral, and negative reviews (must sum to 100)
+2. commonThemes: List 3-6 recurring themes or topics mentioned by users
+3. actionableInsights: List 3-5 specific, actionable insights based on the feedback
+
+${this.translations.respondInLanguage}`;
+    }
+    /**
+     * Generate GitHub-aware improvement suggestions.
+     *
+     * GitHubリポジトリ分析に基づく改善提案を生成。
+     */
+    async generateGitHubImprovements(genai, data, modelName) {
+        try {
+            const prompt = this.buildGitHubImprovementsPrompt(data);
+            const response = await genai.models.generateContent({
+                model: modelName,
+                contents: prompt,
+                config: {
+                    responseMimeType: "application/json",
+                    responseSchema: {
+                        type: genai_1.Type.OBJECT,
+                        properties: {
+                            improvementSummary: { type: genai_1.Type.STRING },
+                            improvements: {
+                                type: genai_1.Type.ARRAY,
+                                items: {
+                                    type: genai_1.Type.OBJECT,
+                                    properties: {
+                                        title: { type: genai_1.Type.STRING },
+                                        description: { type: genai_1.Type.STRING },
+                                        priority: { type: genai_1.Type.STRING },
+                                        category: { type: genai_1.Type.STRING },
+                                        expectedImpact: { type: genai_1.Type.STRING },
+                                        relatedFeature: { type: genai_1.Type.STRING },
+                                        codeReferences: {
+                                            type: genai_1.Type.ARRAY,
+                                            items: {
+                                                type: genai_1.Type.OBJECT,
+                                                properties: {
+                                                    filePath: { type: genai_1.Type.STRING },
+                                                    currentFunctionality: { type: genai_1.Type.STRING },
+                                                    proposedChange: { type: genai_1.Type.STRING },
+                                                    modificationType: { type: genai_1.Type.STRING },
+                                                },
+                                                required: ["filePath", "currentFunctionality", "proposedChange", "modificationType"],
+                                            },
+                                        },
+                                    },
+                                    required: ["title", "description", "priority", "category", "expectedImpact", "codeReferences"],
+                                },
+                            },
+                        },
+                        required: ["improvementSummary", "improvements"],
+                    },
+                },
+            });
+            const text = response.text;
+            if (!text) {
+                throw new Error("No response from AI model");
+            }
+            const inputTokens = response.usageMetadata?.promptTokenCount ?? 0;
+            const outputTokens = response.usageMetadata?.candidatesTokenCount ?? 0;
+            const parsed = JSON.parse(text);
+            const githubRepo = data.githubRepository;
+            return {
+                data: {
+                    repository: githubRepo.repository,
+                    framework: githubRepo.framework,
+                    improvements: parsed.improvements,
+                    improvementSummary: parsed.improvementSummary,
+                    generatedAt: new Date().toISOString(),
+                },
+                inputTokens,
+                outputTokens,
+            };
+        }
+        catch (err) {
+            console.error("Failed to generate GitHub improvements:", err.message);
+            return {
+                data: {
+                    repository: data.githubRepository?.repository ?? "",
+                    framework: data.githubRepository?.framework ?? "",
+                    improvements: [],
+                    improvementSummary: "",
+                    generatedAt: new Date().toISOString(),
+                },
+                inputTokens: 0,
+                outputTokens: 0,
+            };
+        }
+    }
+    /**
+     * Build prompt for GitHub-aware improvements.
+     *
+     * GitHub対応改善提案用のプロンプトを生成。
+     */
+    buildGitHubImprovementsPrompt(data) {
+        const github = data.githubRepository;
+        // フィーチャーをテキストに変換（最大15件に制限）
+        const featuresText = github.features
+            .slice(0, 15)
+            .map((f) => `### ${f.name}\n${f.description}\nFiles: ${f.relatedFiles.join(", ")}`)
+            .join("\n\n");
+        // 全てのファイルパスを収集
+        const allFilePaths = github.features
+            .slice(0, 15)
+            .flatMap((f) => f.relatedFiles)
+            .filter((v, i, a) => a.indexOf(v) === i) // 重複排除
+            .join("\n- ");
+        return `あなたはアプリ開発とマーケティング戦略の専門家です。以下のマーケティングデータとコードベース分析に基づいて、具体的なコード改善提案を行ってください。
+
+## アプリケーション情報
+- Repository: ${github.repository}
+- Framework: ${github.framework}
+- Platforms: ${github.platforms.join(", ")}
+
+## アプリケーション概要
+${github.overview}
+
+## アーキテクチャ
+${github.architecture}
+
+## 機能とファイル
+${featuresText}
+
+## 利用可能なファイルパス
+以下のファイルパスのみをcodeReferencesで使用してください：
+- ${allFilePaths}
+
+## マーケティングデータ
+
+### Google Play Console
+${data.googlePlayConsole ? JSON.stringify(data.googlePlayConsole, null, 2) : "データなし"}
+
+### App Store
+${data.appStore ? JSON.stringify(data.appStore, null, 2) : "データなし"}
+
+### Firebase Analytics
+${data.firebaseAnalytics ? JSON.stringify(data.firebaseAnalytics, null, 2) : "データなし"}
+
+## 指示
+マーケティングデータとコードベース分析に基づいて、5-8個の具体的な改善提案を行ってください。
+
+各提案には必ず以下を含めてください：
+1. **title**: 簡潔なタイトル（50文字以内）
+2. **description**: 詳細な説明（100-200文字）
+3. **priority**: "high", "medium", "low" のいずれか
+4. **category**: "user_acquisition", "retention", "engagement", "monetization", "quality", "development" のいずれか
+5. **expectedImpact**: 期待される効果（具体的な数値目標があれば含める）
+6. **relatedFeature**: 上記の機能名から関連するものを選択
+7. **codeReferences**: 具体的なファイル修正の配列
+   - **filePath**: 上記「利用可能なファイルパス」から選択した実際のパス
+   - **currentFunctionality**: 現在のファイル/機能の説明
+   - **proposedChange**: 提案する具体的な変更内容
+   - **modificationType**: "add"（新規追加）, "modify"（修正）, "refactor"（リファクタリング）, "optimize"（最適化）のいずれか
+
+Important:
+- codeReferences filePath must be selected from the "Available File Paths" list above
+- Focus on code changes that improve marketing metrics (user acquisition, retention, engagement, monetization)
+- Each improvement suggestion should be specific and actionable
+
+${this.translations.respondInLanguage}`;
+    }
+    /**
+     * Generate competitive positioning analysis.
+     *
+     * 競合ポジショニング分析を生成。
+     */
+    async generateCompetitivePositioning(genai, data, modelName) {
+        try {
+            const prompt = this.buildCompetitivePositioningPrompt(data);
+            const response = await genai.models.generateContent({
+                model: modelName,
+                contents: prompt,
+                config: {
+                    responseMimeType: "application/json",
+                    responseSchema: {
+                        type: genai_1.Type.OBJECT,
+                        properties: {
+                            marketPosition: { type: genai_1.Type.STRING },
+                            competitorComparison: {
+                                type: genai_1.Type.ARRAY,
+                                items: {
+                                    type: genai_1.Type.OBJECT,
+                                    properties: {
+                                        competitor: { type: genai_1.Type.STRING },
+                                        ourStrengths: {
+                                            type: genai_1.Type.ARRAY,
+                                            items: { type: genai_1.Type.STRING },
+                                        },
+                                        ourWeaknesses: {
+                                            type: genai_1.Type.ARRAY,
+                                            items: { type: genai_1.Type.STRING },
+                                        },
+                                        battleStrategy: { type: genai_1.Type.STRING },
+                                    },
+                                    required: ["competitor", "ourStrengths", "ourWeaknesses", "battleStrategy"],
+                                },
+                            },
+                            differentiationStrategy: { type: genai_1.Type.STRING },
+                            quickWins: {
+                                type: genai_1.Type.ARRAY,
+                                items: { type: genai_1.Type.STRING },
+                            },
+                        },
+                        required: ["marketPosition", "competitorComparison", "differentiationStrategy", "quickWins"],
+                    },
+                },
+            });
+            const text = response.text;
+            if (!text) {
+                throw new Error("No response from AI model");
+            }
+            const inputTokens = response.usageMetadata?.promptTokenCount ?? 0;
+            const outputTokens = response.usageMetadata?.candidatesTokenCount ?? 0;
+            return {
+                data: JSON.parse(text),
+                inputTokens,
+                outputTokens,
+            };
+        }
+        catch (err) {
+            console.error("Failed to generate competitive positioning:", err.message);
+            return {
+                data: {
+                    marketPosition: "",
+                    competitorComparison: [],
+                    differentiationStrategy: "",
+                    quickWins: [],
+                },
+                inputTokens: 0,
+                outputTokens: 0,
+            };
+        }
+    }
+    /**
+     * Build prompt for competitive positioning analysis.
+     *
+     * 競合ポジショニング分析用のプロンプトを生成。
+     */
+    buildCompetitivePositioningPrompt(data) {
+        const competitors = data.marketResearchData?.competitorAnalysis.competitors || [];
+        const competitorsList = competitors
+            .map((c) => `- ${c.name}: ${c.description}\n  強み: ${c.strengths.join(", ")}\n  弱み: ${c.weaknesses.join(", ")}`)
+            .join("\n");
+        return `あなたは競合分析の専門家です。以下のアプリパフォーマンスデータと市場調査の競合データに基づいて、競合ポジショニング分析を行ってください。
+
+## 当アプリのパフォーマンスデータ
+
+### Google Play Console
+${data.googlePlayConsole ? JSON.stringify(data.googlePlayConsole, null, 2) : "データなし"}
+
+### App Store
+${data.appStore ? JSON.stringify(data.appStore, null, 2) : "データなし"}
+
+### Firebase Analytics
+${data.firebaseAnalytics ? JSON.stringify(data.firebaseAnalytics, null, 2) : "データなし"}
+
+## 市場調査：競合情報
+${competitorsList || "競合データなし"}
+
+## 市場ランドスケープ
+${data.marketResearchData?.competitorAnalysis.marketLandscape || "情報なし"}
+
+## 差別化機会
+${JSON.stringify(data.marketResearchData?.competitorAnalysis.differentiationOpportunities || [], null, 2)}
+
+## 分析指示
+以下の観点で分析してください：
+
+1. **marketPosition**: 現在の市場での位置づけ（2-3文）
+   - 市場シェアの推定
+   - 競合との相対的な強さ
+
+2. **competitorComparison**: 各競合との比較（3-5社）
+   - **competitor**: 競合名
+   - **ourStrengths**: 当アプリの優位点（2-4点）
+   - **ourWeaknesses**: 当アプリの劣位点（2-4点）
+   - **battleStrategy**: この競合に対する具体的な対抗戦略（1-2文）
+
+3. **differentiationStrategy**: 総合的な差別化戦略（3-4文）
+   - 複数の競合に対して有効な差別化ポイント
+
+4. **quickWins**: Quick-win differentiation tactics (3-5 points)
+   - Things that can be executed in 1-2 weeks
+
+${this.translations.respondInLanguage}`;
+    }
+    /**
+     * Generate market opportunity priority analysis.
+     *
+     * 市場機会優先度分析を生成。
+     */
+    async generateMarketOpportunityPriority(genai, data, modelName) {
+        try {
+            const prompt = this.buildMarketOpportunityPriorityPrompt(data);
+            const response = await genai.models.generateContent({
+                model: modelName,
+                contents: prompt,
+                config: {
+                    responseMimeType: "application/json",
+                    responseSchema: {
+                        type: genai_1.Type.OBJECT,
+                        properties: {
+                            prioritizedOpportunities: {
+                                type: genai_1.Type.ARRAY,
+                                items: {
+                                    type: genai_1.Type.OBJECT,
+                                    properties: {
+                                        opportunity: { type: genai_1.Type.STRING },
+                                        fitScore: {
+                                            type: genai_1.Type.STRING,
+                                            enum: ["excellent", "good", "moderate", "poor"],
+                                        },
+                                        fitReason: { type: genai_1.Type.STRING },
+                                        requiredChanges: {
+                                            type: genai_1.Type.ARRAY,
+                                            items: { type: genai_1.Type.STRING },
+                                        },
+                                        estimatedEffort: {
+                                            type: genai_1.Type.STRING,
+                                            enum: ["low", "medium", "high"],
+                                        },
+                                        recommendedAction: { type: genai_1.Type.STRING },
+                                    },
+                                    required: ["opportunity", "fitScore", "fitReason", "requiredChanges", "estimatedEffort", "recommendedAction"],
+                                },
+                            },
+                            strategicRecommendation: { type: genai_1.Type.STRING },
+                        },
+                        required: ["prioritizedOpportunities", "strategicRecommendation"],
+                    },
+                },
+            });
+            const text = response.text;
+            if (!text) {
+                throw new Error("No response from AI model");
+            }
+            const inputTokens = response.usageMetadata?.promptTokenCount ?? 0;
+            const outputTokens = response.usageMetadata?.candidatesTokenCount ?? 0;
+            return {
+                data: JSON.parse(text),
+                inputTokens,
+                outputTokens,
+            };
+        }
+        catch (err) {
+            console.error("Failed to generate market opportunity priority:", err.message);
+            return {
+                data: {
+                    prioritizedOpportunities: [],
+                    strategicRecommendation: "",
+                },
+                inputTokens: 0,
+                outputTokens: 0,
+            };
+        }
+    }
+    /**
+     * Build prompt for market opportunity priority analysis.
+     *
+     * 市場機会優先度分析用のプロンプトを生成。
+     */
+    buildMarketOpportunityPriorityPrompt(data) {
+        const opportunities = data.marketResearchData?.businessOpportunities || [];
+        const opportunitiesList = opportunities
+            .map((o) => `- **${o.title}** (${o.type}, インパクト: ${o.potentialImpact}, 時間枠: ${o.timeframe})\n  ${o.description}\n  要件: ${o.requirements.join(", ")}\n  リスク: ${o.risks.join(", ")}`)
+            .join("\n\n");
+        const marketGaps = data.marketResearchData?.competitorAnalysis.marketGaps || [];
+        return `あなたは戦略コンサルタントです。以下のアプリパフォーマンスデータと市場機会データに基づいて、機会の優先順位付けを行ってください。
+
+## 当アプリのパフォーマンスデータ
+
+### Google Play Console
+${data.googlePlayConsole ? JSON.stringify(data.googlePlayConsole, null, 2) : "データなし"}
+
+### App Store
+${data.appStore ? JSON.stringify(data.appStore, null, 2) : "データなし"}
+
+### Firebase Analytics
+${data.firebaseAnalytics ? JSON.stringify(data.firebaseAnalytics, null, 2) : "データなし"}
+
+## 市場調査：ビジネス機会
+${opportunitiesList || "機会データなし"}
+
+## 市場調査：市場ギャップ
+${JSON.stringify(marketGaps, null, 2)}
+
+## 需要予測サマリー
+${data.marketResearch?.demandForecast.summary || "情報なし"}
+
+## 分析指示
+各ビジネス機会と市場ギャップを、当アプリの現状と照らし合わせて優先順位付けしてください。
+
+1. **prioritizedOpportunities**: 優先順位付けされた機会リスト（5-8件）
+   - **opportunity**: 機会名（市場調査データから）
+   - **fitScore**: 当アプリとの適合度
+     - "excellent": 既存機能・リソースで即座に対応可能
+     - "good": 小規模な追加開発で対応可能
+     - "moderate": 中規模の開発・投資が必要
+     - "poor": 大規模な方向転換が必要
+   - **fitReason**: 適合度の理由（1-2文）
+   - **requiredChanges**: 必要な変更・施策（2-4点）
+   - **estimatedEffort**: 実装工数の見積もり（low/medium/high）
+   - **recommendedAction**: 推奨アクション（1文）
+
+2. **strategicRecommendation**: Strategic recommendations (3-4 sentences)
+   - Priority opportunities to pursue
+   - Resource allocation proposals
+
+${this.translations.respondInLanguage}`;
+    }
+}
+exports.AnalyzeMarketingData = AnalyzeMarketingData;
+module.exports = (regions, options, data) => new AnalyzeMarketingData(options).build(regions);
+// Export class for testing
+module.exports.AnalyzeMarketingData = AnalyzeMarketingData;
+//# sourceMappingURL=analyze_marketing_data.js.map