Energy Transition Resilience MCP

Energy transition resilience analysis via the Model Context Protocol, combining 8 live data sources into a unified multi-factor assessment. This MCP server is built for energy analysts, infrastructure investors, policymakers, and AI agents that need structured, quantitative answers about how ready a region is to weather the transition away from fossil fuels. It produces scored assessments across EV readiness, fossil fuel dependency, air quality, climate vulnerability, and economic price shock risk — all from a single tool call.

Connect it to Claude, Cursor, Windsurf, or any MCP-compatible client and ask questions like "How resilient is Germany's energy system?" or "What is the price shock risk for USD-denominated energy imports?" and receive structured JSON answers backed by live government and open-data sources.

⬇️ What data can you access?

Why use Energy Transition Resilience MCP?

Manually assessing a region's energy transition readiness requires pulling data from eight different government portals, harmonizing units, building scoring models, and refreshing everything when indicators change. That process takes days for an analyst and weeks to automate from scratch.

This MCP replaces that pipeline with a single tool call. You get scored, structured results — ready for AI agent consumption, dashboard injection, or direct analysis — in seconds.

Beyond raw convenience, the server runs on the Apify platform, which means:

• Scheduling — run weekly resilience scans for portfolio regions and store results automatically
• API access — trigger assessments from Python, TypeScript, or any HTTP client without managing infrastructure
• Pay-per-call pricing — no subscription, no idle costs; pay only for tool calls made
• Monitoring — get alerts when runs fail or score thresholds are breached
• Integrations — push results directly to Google Sheets, Slack, Zapier, or Make workflows

❓ MCP tools

Features

• 8-source parallel data fetching — all Apify actor calls run concurrently via Promise.all, minimizing latency
• Haversine proximity edge-building — EV stations and air quality monitors within 100 km are linked in the network graph, with edge weights decaying by distance
• Coefficient of variation volatility scoring — FRED, World Bank, and ECB data series are analyzed for mean/variance ratios, with currency volatility weighted 2x in the price shock model
• 5-dimension weighted composite scoring — EV readiness (25%), fossil dependency (20%), air quality (15%), climate resilience (20%), economic stability (20%)
• Transition grade A-F — overall score maps to a letter grade: A (80+), B (65-79), C (50-64), D (35-49), F (<35)
• Risk-level classification — four tiers (low, moderate, high, critical) on all scored dimensions
• Fossil fuel dependency index — combines license volume, active ratio, and type diversity into a 0-100 index; higher = more dependent
• EV readiness sub-scoring — density per 100,000 population (max 30 pts), fast-charger ratio (30 pts), average kW (20 pts), total station count (20 pts)
• Air quality correlation score — PM2.5 normalized to 35 µg/m³ (60% weight) + PM10 normalized to 50 µg/m³ (40% weight)
• Climate vulnerability score — Extreme alerts (15 pts each), Severe alerts (8 pts each), Moderate (3 pts each), air quality degradation bonus (up to 20 pts)
• Eurostat integration — EU energy balance dataset (nrg_bal_c) included in network mapping and full resilience reports
• SSE transport — runs as a persistent Server-Sent Events MCP server on Apify's Standby infrastructure; no cold-start latency after initialization
• Structured JSON output — every tool returns typed results matching defined TypeScript interfaces, safe for programmatic consumption

Use cases for energy transition resilience analysis

Municipal energy planning and infrastructure investment

City planners and regional governments need to prioritize infrastructure spending — whether that means EV charging expansion, grid modernization, or fossil fuel asset retirement. The compute_transition_readiness_score tool produces a single scored assessment across all five dimensions with actionable recommendations ("Expand fast chargers >50 kW", "Diversify energy sources to reduce price shock vulnerability"), giving budget committees a defensible, data-backed starting point.

Energy portfolio risk assessment

Asset managers and infrastructure funds holding exposure to fossil fuel producers, utilities, or clean-energy developers need quantified transition risk per region. The compute_fossil_fuel_dependency and detect_energy_price_shock_risk tools provide the dependency index and volatility scores needed to weight holdings by transition-risk tier. Running these tools weekly via scheduled Apify runs keeps portfolio risk scores current without analyst intervention.

Climate adaptation and resilience strategy

Insurers, development banks, and NGOs planning climate adaptation investments need to know which regions face the greatest combination of climate hazard and energy infrastructure vulnerability. The assess_climate_vulnerability tool aggregates NOAA weather severity data and OpenAQ air quality degradation into a single vulnerability score, with the top 5 threat types ranked by frequency.

EV network expansion planning

Charging network operators evaluating new market entry need station density, coverage classification (insufficient / developing / adequate / excellent), and fast-charger penetration by region before committing capital. The assess_ev_infrastructure_readiness tool pulls live Open Charge Map data and scores density per 100,000 population, fast-charger ratio, and average kW output in a single call.

Air quality policy and public health analysis

Environmental regulators and public health agencies correlating energy mix with pollution outcomes need normalized pollutant levels tied to fossil fuel exposure proxies. The analyze_air_quality_correlation tool computes a correlation score linking PM2.5 and PM10 averages to the region's energy profile, supporting the regulatory case for accelerated transition.

AI agent energy research workflows

LLM-powered research agents in Claude, Cursor, or custom pipelines can call any of the 8 tools directly via MCP to answer natural language questions about energy resilience. The generate_energy_resilience_report tool returns a structured executive summary plus all sub-dimension scores in a single response — ideal for feeding into summarization or report-generation pipelines.

How to connect this MCP server

Step 1: Get your Apify API token

Sign up at apify.com (free tier includes $5/month in credits). Copy your API token from Account Settings.

Step 2: Connect your MCP client

Claude Desktop — add to claude_desktop_config.json:

{
  "mcpServers": {
    "energy-transition-resilience": {
      "url": "https://energy-transition-resilience-mcp.apify.actor/sse",
      "headers": {
        "Authorization": "Bearer YOUR_APIFY_TOKEN"
      }
    }
  }
}

Cursor / Windsurf / Cline — use the same SSE URL in your MCP settings panel.

Step 3: Call a tool

Ask your AI client: "Use the energy transition MCP to compute transition readiness for Germany." The client will call compute_transition_readiness_score with region: "DE" and return a scored, structured result.

Step 4: Download results

Full JSON responses are returned inline via MCP. For batch processing or scheduled runs, results are also stored in the Apify dataset for the run.

Input parameters

The MCP server requires only one configuration parameter:

All analytical parameters (region, data sources, max results) are passed as tool arguments at call time — not as actor input.

Tool arguments reference

⬆️ Output example

Response from generate_energy_resilience_report with region: "DE":

{
  "region": "DE",
  "reportDate": "2026-03-21",
  "executiveSummary": "Energy Transition Resilience Report for DE. Overall resilience score: 62/100 (moderate risk). EV readiness: 58/100 (adequate). Fossil dependency: 34/100 (moderate). Transition grade: B.",
  "evInfrastructure": {
    "region": "DE",
    "stationCount": 87,
    "densityPerCapita": 0.00087,
    "readinessScore": 58,
    "chargerTypes": {
      "CCS (Type 2)": 64,
      "CHAdeMO": 23,
      "Type 2 (Mennekes)": 102
    },
    "coverage": "adequate",
    "stations": [],
    "timestamp": "2026-03-21T09:14:22.000Z"
  },
  "fossilFuelDependency": {
    "region": "DE",
    "licenseCount": 31,
    "dependencyIndex": 34,
    "licenseTypes": {
      "Exploration": 12,
      "Production": 19
    },
    "riskLevel": "moderate",
    "licenses": [],
    "timestamp": "2026-03-21T09:14:22.000Z"
  },
  "airQualityCorrelation": {
    "region": "DE",
    "measurementCount": 87,
    "averagePM25": 14.2,
    "averagePM10": 22.8,
    "correlationScore": 42,
    "pollutants": {
      "pm25": 14.2,
      "pm10": 22.8,
      "no2": 31.4,
      "o3": 67.1
    },
    "measurements": [],
    "timestamp": "2026-03-21T09:14:22.000Z"
  },
  "climateVulnerability": {
    "region": "DE",
    "alertCount": 14,
    "severityDistribution": {
      "Severe": 4,
      "Moderate": 9,
      "Minor": 1
    },
    "vulnerabilityScore": 61,
    "topThreats": [
      "Winter Storm Warning",
      "Flood Advisory",
      "High Wind Warning",
      "Freezing Fog Advisory",
      "Dense Fog Advisory"
    ],
    "alerts": [],
    "timestamp": "2026-03-21T09:14:22.000Z"
  },
  "economicIndicators": {
    "indicators": [],
    "volatilityIndex": 38,
    "shockProbability": 0.19,
    "riskFactors": [
      "FRED indicator: Crude Oil Prices: West Texas Intermediate",
      "FRED indicator: Henry Hub Natural Gas Spot Price",
      "High coefficient of variation in FRED data: 31.4%"
    ],
    "riskLevel": "moderate",
    "timestamp": "2026-03-21T09:14:22.000Z"
  },
  "transitionReadiness": {
    "region": "DE",
    "overallScore": 62,
    "dimensions": {
      "evReadiness": 58,
      "fossilDependency": 66,
      "airQuality": 58,
      "climateResilience": 39,
      "economicStability": 62
    },
    "grade": "B",
    "recommendations": [
      "Strengthen climate adaptation measures for energy infrastructure"
    ],
    "timestamp": "2026-03-21T09:14:22.000Z"
  },
  "overallResilienceScore": 62,
  "riskLevel": "moderate",
  "timestamp": "2026-03-21T09:14:22.000Z"
}

Output fields

Top-level (generate_energy_resilience_report)

EV infrastructure sub-report

Fossil fuel dependency sub-report

Air quality correlation sub-report

Climate vulnerability sub-report

Energy price shock risk sub-report

Transition readiness score

How the scoring algorithms work

EV infrastructure readiness scoring

The assessEVInfrastructureReadiness function scores charging density and quality on a 100-point scale. Points are awarded in four buckets: station density per 100,000 population (up to 30 points), fast-charger ratio where "fast" means 50 kW or above (30 points), average kW per station normalized to 150 kW max (20 points), and total station count capped at 100 stations for full marks (20 points). The final readiness score maps to four coverage tiers: insufficient (<25), developing (25-49), adequate (50-74), and excellent (75+).

Fossil fuel dependency index

The computeFossilFuelDependency function builds a 0-100 index from three signals: license volume normalized to 50 licenses for full marks (40 points), the ratio of active/extant/current licenses to total licenses (35 points), and license type variety normalized to 5 distinct types (25 points). The active-license check matches status strings containing "active", "extant", or "current" to count operating assets. Higher scores indicate greater dependency and higher transition risk.

Energy price shock risk with coefficient of variation

The detectEnergyPriceShockRisk function computes coefficient of variation (CV = standard deviation / mean) independently for FRED, World Bank, and ECB data series. Currency volatility (ECB) is weighted 2x because exchange rate instability amplifies energy import costs. The three CVs are averaged into avgVolatility, which maps to a volatilityIndex (scaled ×200, capped at 100) and a shockProbability (scaled ×300, capped at 1.0). Risk tiers: low (<25), moderate (25-49), high (50-74), critical (75+).

Climate vulnerability scoring

The assessClimateVulnerability function scores weather alert severity distribution: Extreme alerts contribute 15 points each, Severe 8 points, Moderate 3 points. Air quality degradation adds bonus points: PM2.5 above 25 µg/m³ adds 0.5 per measurement, PM10 above 50 µg/m³ adds 0.3, O3 above 100 µg/m³ adds 0.4 — the total is capped at 20. Alert volume also contributes: total alerts divided by 10, capped at 1, scaled to 15 points. All contributions sum and are capped at 100.

Air quality correlation score

PM2.5 averages are normalized to the 35 µg/m³ warning threshold (60% weight) and PM10 to 50 µg/m³ (40% weight). The combined correlation score reflects how far measured pollution levels exceed safe thresholds. A score of 100 means pollution levels fully explain fossil fuel exposure; 0 means ambient pollution is below warning thresholds.

Transition readiness composite

The weighted composite applies: EV readiness (25%) + inverted fossil dependency (20%) + inverted air quality score (15%) + inverted climate vulnerability (20%) + inverted economic volatility (20%). Inversion ensures that all five dimensions contribute positively when conditions are good: lower fossil dependency, lower pollution, lower climate vulnerability, and lower price volatility all raise the composite score. Letter grades: A (80+), B (65-79), C (50-64), D (35-49), F (<35).

Energy transition network graph

The buildEnergyNetwork function creates typed nodes for each data source (ev_station, oil_license, air_quality, weather_event, economic_indicator, currency, stat_series) and builds three categories of edges:

• EV-to-air-quality edges: haversine distance between station coordinates and monitor coordinates; edges created for pairs within 100 km, weighted by 1 - (distance / 100)
• Oil-to-economic edges: all oil license nodes linked to all economic indicator nodes with weight 0.5 (representing economic impact channel)
• Weather-to-EV edges: all weather event nodes linked to the 5 nearest EV station nodes, weighted 1.0 for Extreme, 0.75 for Severe, 0.4 otherwise

How much does it cost to use this MCP?

Every tool call uses pay-per-event pricing at $0.045 per call. There is no subscription, no minimum commitment, and no charge for idle time.

Apify's free tier includes $5 of monthly platform credits. At $0.045 per call, that covers 111 individual tool calls at no cost.

Using the API

Python

from apify_client import ApifyClient

client = ApifyClient("YOUR_API_TOKEN")

# Call compute_transition_readiness_score via the actor run API
run = client.actor("ryanclinton/energy-transition-resilience-mcp").call(run_input={
    "port": 3039
})

# For direct MCP tool calls, use the SSE endpoint from your MCP client
# Results from scheduled runs are available in the dataset:
for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(f"Region: {item.get('region')} | Score: {item.get('overallResilienceScore')} | Grade: {item.get('grade')}")

JavaScript

import { ApifyClient } from "apify-client";

const client = new ApifyClient({ token: "YOUR_API_TOKEN" });

const run = await client.actor("ryanclinton/energy-transition-resilience-mcp").call({
    port: 3039
});

const { items } = await client.dataset(run.defaultDatasetId).listItems();
for (const item of items) {
    console.log(`Region: ${item.region} | Score: ${item.overallResilienceScore} | Risk: ${item.riskLevel}`);
}

Direct MCP call (cURL)

# Call generate_energy_resilience_report directly via the MCP HTTP endpoint
curl -X POST "https://energy-transition-resilience-mcp.apify.actor/messages?sessionId=SESSION_ID" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_APIFY_TOKEN" \
  -d '{
    "jsonrpc": "2.0",
    "method": "tools/call",
    "params": {
      "name": "generate_energy_resilience_report",
      "arguments": {
        "region": "DE",
        "sources": ["ev_infra", "fossil", "air_quality", "weather", "us_econ", "global_econ"],
        "maxResults": 50
      }
    },
    "id": 1
  }'

# Open an SSE session first to get a sessionId:
curl -N "https://energy-transition-resilience-mcp.apify.actor/sse" \
  -H "Authorization: Bearer YOUR_APIFY_TOKEN"

Tips for best results

1. Start with generate_energy_resilience_report for a new region before drilling into individual dimensions. The full report costs the same as any single-dimension call ($0.045) and gives you the complete picture, including the A-F grade and specific recommendations.

2. Use maxResults: 100 for EV readiness assessments in countries with large charging networks (UK, DE, FR, NL). The default of 50 is sufficient for smaller markets. Higher counts improve the density and fast-charger-ratio calculations.

3. Select specific sources when you only need a subset. If you are only modeling economic price shock risk, call detect_energy_price_shock_risk directly rather than the full network map — this is faster and costs the same.

4. Use ISO 2-letter country codes for consistency. Region accepts any string, but using ISO codes (DE, FR, GB, US, JP) ensures reliable matching across Open Charge Map, OpenAQ, and World Bank APIs.

5. Schedule weekly resilience reports for portfolio regions using Apify's scheduling feature. Store results in the dataset and track overallResilienceScore over time to detect deteriorating regions before they become headline risk.

6. Combine with the Energy Transition Intelligence MCP for US-focused analysis. The Resilience MCP provides the multi-country comparative view; the Intelligence MCP adds EIA grid stress and US carbon intensity signals for deeper US coverage.

7. For AI agent pipelines, call compute_transition_readiness_score first for the scored summary, then call specific sub-tools (e.g., analyze_air_quality_correlation) only if the agent determines a dimension needs deeper investigation. This minimizes cost in automated workflows.

Combine with other Apify actors

Limitations

• Fossil fuel license data is UK-only. The NSTA data source covers North Sea licenses exclusively. The compute_fossil_fuel_dependency tool will return limited or zero results for non-UK regions. For other geographies, only the dependencyIndex derived from available records is meaningful.
• NOAA weather alerts are US-centric. The assess_climate_vulnerability tool works globally via OpenAQ air quality data, but weather severity distribution reflects NOAA's US alert system. Non-US regions will show low alert counts that underestimate actual climate hazards.
• EV station data quality varies by country. Open Charge Map coverage is denser in Europe, the US, and Australia. Remote regions or emerging markets may return insufficient station counts, causing readinessScore to reflect data scarcity rather than true infrastructure gaps.
• Economic indicator freshness varies. FRED data updates daily to quarterly depending on the series. World Bank indicators may be 1-2 years behind. ECB rates update daily. The volatilityIndex reflects historical variance, not forward-looking projections.
• Scores are normalized to the data available. If fewer than the requested maxResults records are returned by a source, scores are computed on the available subset. Smaller datasets produce less reliable coefficients of variation.
• No historical trend tracking. Each tool call returns a point-in-time assessment. Trend analysis requires scheduling repeated runs and comparing overallResilienceScore values over time via the Apify dataset.
• Concurrent actor calls consume Apify credits. The generate_energy_resilience_report and map_energy_transition_network tools dispatch up to 8 parallel actor calls. These sub-actor calls are charged separately against your Apify account in addition to the $0.045 MCP tool call fee.
• No custom weighting. The 5-dimension composite weights (EV: 25%, fossil: 20%, air: 15%, climate: 20%, economic: 20%) are fixed in code. If your use case requires different weighting, call individual sub-tools and apply custom weights in post-processing.

Integrations

• Apify API — trigger resilience assessments and retrieve results programmatically from any language
• Webhooks — get notified when scheduled resilience reports complete, enabling downstream alerts or data pushes
• Zapier — push resilience scores to spreadsheets, Slack channels, or CRM systems when runs complete
• Make — build no-code workflows that trigger region assessments on a schedule and route results to dashboards
• Google Sheets — export transition readiness scores and dimension breakdowns for portfolio monitoring
• LangChain / LlamaIndex — use this MCP server as a tool provider in RAG pipelines or autonomous energy research agents

Troubleshooting

• Tool returns empty or null sub-reports — this is expected when the source data returns zero records. The NSTA fossil fuel source is UK-only; non-UK regions return an empty license set, so fossilFuelDependency will be null in the full report. Use compute_fossil_fuel_dependency only for UK regions.

• Low readiness scores despite known infrastructure — verify the region argument uses the correct format expected by each underlying API. Open Charge Map uses country codes like GB or US. OpenAQ also uses 2-letter ISO codes. Mismatched codes (e.g., UK vs GB) cause data fetches to return zero results, driving scores down artificially.

• Run timeout on full resilience reports — the generate_energy_resilience_report tool fires 8 parallel actor calls with a 120-second timeout each. On Apify's free tier, concurrent actor slots may be limited. Reduce maxResults to 25 or select a subset of sources to speed up execution.

• MCP client shows connection errors — confirm you are using the SSE URL (/sse endpoint) and not the actor run API URL. The MCP session starts with a GET to /sse; tool calls go to /messages?sessionId=SESSION_ID. Using the wrong endpoint returns a 400 error.

• shockProbability is always 0 — this happens when only one data point is returned per source (CV requires at least two values). Increase maxResults or broaden searchText for the FRED source to ensure multiple indicator series are returned.

Responsible use

• This MCP accesses only publicly available data from government registries and open-data platforms.
• Open Charge Map, NSTA, OpenAQ, NOAA, FRED, World Bank, ECB, and Eurostat data is available under their respective open-access terms.
• Do not use resilience scores to make investment decisions without validation against primary source data and domain expert review.
• For guidance on web scraping legality, see Apify's guide.

FAQ

How do I compute energy transition readiness for a specific country? Call compute_transition_readiness_score with the ISO 2-letter country code as region (e.g., "FR" for France, "DE" for Germany, "US" for the United States). The tool fetches data from all 6 sources simultaneously and returns the 5-dimension weighted composite with a letter grade.

How many tool calls can I make per month for free? Apify's free tier includes $5 of monthly credits. At $0.045 per tool call, that covers 111 calls per month at no cost. Paid plans start at $49/month with substantially higher credit allocations.

What regions and countries does this MCP cover? EV infrastructure (Open Charge Map) and air quality (OpenAQ) are global. World Bank and ECB cover all major economies. FRED covers US economic indicators. Fossil fuel license data (NSTA) covers the UK North Sea only. Eurostat covers EU member states. Climate vulnerability (NOAA weather) is primarily US-focused.

How accurate is the transition readiness score? The score is a normalized composite of available data, not a certified benchmark. It is calibrated to produce meaningful relative rankings between regions — a region scoring 75 is genuinely better positioned than one scoring 35. Absolute values should be interpreted as directional indicators, not precise measurements. The letter grade (A-F) is the most robust output for stakeholder communication.

How is this different from the Energy Transition Intelligence MCP? The Energy Transition Intelligence MCP focuses on US energy markets using EIA grid generation and carbon intensity data. This Resilience MCP takes a multi-country view with 8 data sources including EU statistics (Eurostat), global development indicators (World Bank), and normalized air quality correlation, making it better suited for international comparative analysis.

Can I use this MCP in an AI agent pipeline? Yes. The server runs as a persistent SSE MCP endpoint on Apify's Standby infrastructure. Any MCP-compatible client — Claude, Cursor, Windsurf, Cline, or a custom LangChain agent — can connect and call tools. Results return as structured JSON, safe for downstream parsing.

How fresh is the data? Each tool call fetches data live from the underlying sources at query time. Air quality and weather data are near real-time (minutes to hours). FRED and ECB rates update daily. World Bank development indicators are updated annually. Eurostat energy balance data is updated quarterly.

Can I schedule resilience monitoring for multiple regions? Yes. Use Apify's built-in scheduler to run the actor on any interval (daily, weekly, monthly). Combine with the Apify API or webhooks to route results to a Google Sheet or dashboard. Results for each run are stored in the actor's dataset.

Is it legal to use government and open data this way? Yes. All 8 data sources — Open Charge Map, NSTA, OpenAQ, NOAA, FRED, World Bank, ECB, and Eurostat — publish data under open-access or public-domain terms. This MCP is a structured client for these APIs. See Apify's guide on web scraping legality.

What happens if one data source is unavailable? Each source is called with a 120-second timeout. If a source returns an error, the MCP logs the error and continues processing available sources. The affected sub-report will return null or an empty result in the final output. The executiveSummary will reflect only the dimensions where data was successfully retrieved.

Can I weight the 5 dimensions differently for my use case? The composite weights (EV 25%, fossil 20%, air quality 15%, climate 20%, economic 20%) are fixed. To apply custom weights, call the 5 individual sub-tools (assess_ev_infrastructure_readiness, compute_fossil_fuel_dependency, analyze_air_quality_correlation, assess_climate_vulnerability, detect_energy_price_shock_risk) and apply your own weighting to the scores returned.

How does the network graph work? The map_energy_transition_network tool builds a graph with typed nodes for each data record (ev_station, oil_license, air_quality, weather_event, economic_indicator, currency, stat_series). Edges represent three relationship types: EV-to-air-quality proximity (haversine <100 km), oil-license-to-economic-indicator impact, and weather-event-to-EV-station climate vulnerability. The result includes full node and edge arrays plus a summary string with counts by node type.

Help us improve

If you encounter issues, enable run sharing to help us debug faster:

1. Go to Account Settings > Privacy
2. Enable Share runs with public Actor creators

This lets us see your run details when something goes wrong. Your data is only visible to the actor developer, not publicly.

Support

Found a bug or have a feature request? Open an issue in the Issues tab on this actor's page. For custom integrations or enterprise energy analysis workflows, reach out through the Apify platform.