Corrently

@Corrently what's the hourly solar forecast for Berlin (52.52, 13.40) over the next 24 hours?

Based on the location coordinates, the forecast shows peak generation around 12:00-14:00 UTC tomorrow with approximately 4.2 kWh expected output per installed kWp. Morning hours (06:00-09:00) show gradual ramp-up from 0.3 to 2.1 kWh/kWp, while evening decline begins after 16:00. Cloud cover between 10:00-11:00 introduces an estimated 18% loss factor. Total daily yield projects to 22.7 kWh per kWp installed capacity. This data reflects typical May conditions for the region.

This example demonstrates the PV Generation Forecast tool's ability to provide hourly granularity for energy planning. Requires latitude/longitude coordinates. Forecasts depend on weather API availability and are estimates—actual generation varies with panel orientation, shading, and real-time conditions. Useful for scheduling battery charging or load-shifting decisions.

@Corrently log a new meter reading: 847.3 kWh consumed as of today at 14:00 for meter ID home-main

CO₂ meter reading recorded successfully. Meter ID 'home-main' updated with 847.3 kWh consumption timestamped 2024-01-15 14:00 UTC. This reading calculates to approximately 338.9 kg CO₂ equivalent based on current German grid mix (400g CO₂/kWh average). Your cumulative tracked emissions for this meter now total 1,247 kg CO₂ since monitoring began. Next recommended reading: 2024-01-22 to maintain weekly tracking cadence.

This example shows the CO₂ Meter Update tool creating an emissions tracking entry. The MCP writes data to Corrently's backend—this is a side-effect action, not just a read. Requires a valid meter identifier in your Corrently account. CO₂ calculations use regional grid carbon intensity, which fluctuates. Best used for consistent interval logging to build meaningful trend data.

@Corrently compare my last three meter readings and estimate if I'm on track to reduce emissions 15% this quarter

Analyzing your recent readings: Dec 15 (782 kWh), Jan 1 (823 kWh), Jan 15 (847 kWh). Your consumption trend shows a 4.2% increase over the 31-day period, translating to roughly 312 kg CO₂/month at current grid intensity. To achieve a 15% quarterly reduction from your Q4 baseline (2,340 kg CO₂), you'd need to average 665 kg/month—you're currently 47 kg/month over target. Recommendation: shift 12-15% of daytime load to solar peak hours (11:00-15:00) when grid carbon intensity drops 30-40%, or reduce baseline consumption by approximately 95 kWh/month.

This synthesis example combines the MCP's meter data with AI reasoning to provide actionable insights. The MCP supplies raw readings; the AI calculates trends and generates recommendations. Accuracy depends on having multiple logged readings and understanding your regional grid's carbon profile. Most valuable when paired with the solar forecast tool to identify optimal consumption windows.

Track emissions when your team runs multi-region cloud workloads

A 6-person engineering team at a B2B SaaS company needs monthly carbon reports for their SOC 2 audit trail. They pull AWS CloudWatch metrics into Switchy, feed consumption data to the CO₂ Meter Update tool, and generate timestamped emissions logs without leaving their workspace. This MCP works if your electricity usage is already instrumented and you need a lightweight API to turn kWh into CO₂e. It breaks down if you're tracking Scope 3 emissions across vendors or need multi-country grid factors—Corrently focuses on German grid data by default. If your compliance requirement is "prove we measured it" rather than "model every supplier," this is the fastest path from meter reading to audit-ready log.

When to run batch jobs if your office has rooftop panels

A 3-person hardware startup shares a coworking space with solar panels and wants to schedule their nightly CI/CD runs during peak generation hours. They use the PV Generation Forecast tool in Switchy to pull hourly output estimates, then shift their GitHub Actions triggers to match the forecast windows. This MCP is the right call if you have a fixed location, known panel specs, and predictable compute workloads. It's not a fit if your jobs are latency-sensitive or your team is distributed across time zones—the forecast is location-specific and assumes you can defer work by 6-12 hours. If your build queue can wait for sunshine and you want a one-line API call instead of a weather dashboard, Corrently delivers the forecast without the ceremony.

Pass through compute costs when you bill by resource usage

A 4-person data consultancy runs client models on dedicated hardware and wants to itemize energy costs on invoices. They log each job's runtime and kWh draw, push readings to the CO₂ Meter Update tool, and export the emissions data alongside compute hours in their billing CSV. This MCP works if your pricing model already includes resource metering and your clients care about carbon accounting. It's overkill if you're billing flat monthly retainers or your workloads are too small to justify per-job tracking—setup overhead is 20-30 minutes per client meter. If you're already tracking usage and need a defensible emissions number to attach to the invoice line, Corrently turns your meter data into a billable audit trail.