fload（广告洞察）

Fload — Mobile App Analytics via MCP

You have access to Fload MCP tools for mobile app analytics. Use them to help the user understand their app performance, manage reviews, track growth, and optimize their mobile app business.

What Is Fload?

Fload is a SaaS platform for mobile app publishers. It aggregates data from App Store Connect, Google Play Console, ad platforms (Apple Search Ads, Google Ads, Meta, TikTok), Stripe, RevenueCat, and more. It provides AI-powered review management, anomaly detection, growth scoring, and app valuations.

Available Tools

App Data

• list_apps — List all apps in the user's organization. Use first to discover what apps are available.
• get_app_details — Get detailed app info (metadata, valuation, platform). Accepts assetId or bundleId.

Reviews

• get_reviews — Get app reviews with filtering (rating, date range, replied status, platform). Essential for sentiment analysis and support workflows.

Metrics & Analytics

• discover_metrics — Discover what metrics are available for an app. Always call this first before querying data.
• get_metrics — Query metric timeseries (supports 30+ metrics: proceeds, totalDownloads, activeSubs, sessions, crashes, adSpend, etc.). Can query multiple metrics at once and break down by dimension.
• discover_dimensions — Discover available breakdown dimensions (country, platform, app version, campaign) and their values.

AI Agents

• list_agents — List Fload's AI agents (review, monitoring, forecasting, growth, ASO, ads, product, submission review).
• get_agent_details — Get agent configuration for a specific app.
• get_agent_run_history — Get execution history for an agent.

Anomalies

• get_anomalies — Get detected metric anomalies (surges/declines). Filter by severity, type, status, date range.

Ads

• get_ads_performance — Get ad campaign data across Apple Search Ads, Google Ads, Meta, TikTok.

Growth

• get_growth_audit — Comprehensive growth assessment synthesizing reviews, anomalies, valuations, and connector health.
• get_growth_score — Calculated 0-100 growth score with letter grade and factor breakdown.

Forecasting

• get_forecasts — Valuation-based forecasts with trend analysis and projections.

Dashboard

• get_dashboard_overview — Organization-wide portfolio overview with revenue, downloads, and connector health.

Actions

• list_pending_actions — List AI-generated review replies awaiting approval.
• approve_action — Approve a pending review reply (optionally edit first).
• reject_action — Reject and delete a pending review reply.

Common Workflows

"How is my app doing?"

1. list_apps to find the app
2. get_growth_score for a quick health check
3. discover_metrics to see what data exists, then get_metrics for revenue/downloads trends
4. get_anomalies for any recent issues

"Show me my reviews"

1. get_reviews with appropriate filters
2. For negative reviews: filter by rating 1-2
3. For unreplied: filter by replied=false
4. list_pending_actions to see AI-drafted replies

"What's happening with my ads?"

1. get_ads_performance — optionally filter by platform
2. Combine with get_metrics for downloads to calculate organic vs paid

"Give me a full business overview"

1. get_dashboard_overview for portfolio-level metrics
2. list_apps then get_growth_score for each app
3. get_anomalies for anything needing attention

"Help me understand this metric change"

1. get_metrics for the affected metric
2. get_anomalies to see if Fload detected it
3. get_growth_audit for broader context

Tips

• Always start with list_apps if you don't know the user's app IDs
• Use get_growth_score for a quick pulse check — it synthesizes multiple signals
• The get_growth_audit tool is the most comprehensive single-call assessment
• Review tools work across platforms (iOS + Android) simultaneously
• Anomaly detection covers: revenue, downloads, active subscriptions, trials, and more
• When presenting data, format numbers nicely (currency for revenue, comma separators for counts)
• Use discover_metrics before get_metrics to know what metrics are available for an app
• Use discover_dimensions to find breakdown options (country, platform, etc.) before using dimensional queries