Solcast
Solar resource assessment and forecasting data for irradiance and PV power, globally.
Verdict
Common use cases
- Model solar farm output before site approval
- Validate PV system performance against actuals
- Pull weather data for energy audits
- Compare irradiance across candidate locations
- Update panel specs when retrofitting arrays
Integration
- Vendor
- Solcast
- Category
- developer-tools
- Auth
- API_KEY
- Tools
- 6
- Composio slug
solcast
Tools
- Create PV Power Resource
Tool to create a new PV power resource for Advanced PV Power forecasting. Use when you need to add a new solar installation site with comprehensive technical specifications including capacity, orientation, tracking parameters, and derating
- Get Historic Radiation And Weather
Tool to get historical irradiance and weather estimated actuals for a requested location. Use when you need historical solar radiation and weather data from 2007-01-01 to 7 days ago, for up to 31 days at a time.
- Get PV Power Resource
Tool to retrieve a specific PV power resource by its unique identifier. Use when you need to fetch complete resource configuration including capacity, orientation, tracking, and derating parameters.
- Get Resources PV Power Sites
Tool to list PV power resources available to authenticated user. Use when you need to retrieve the list of PV power sites with optional filtering by entitlement level.
- Patch PV Power Resource
Tool to partially update an existing PV power resource's technical specifications. Use when you need to modify specific resource configuration fields without replacing the entire resource.
- Update PV Power Resource
Tool to update an existing PV power resource with comprehensive technical specifications for solar installation site modeling. Use when you need to modify resource configuration including capacity, orientation, tracking parameters, derating
Setup
Setup guide
- 11. Open your Switchy workspace and navigate to Settings > Integrations. 2. Find Solcast in the MCP catalog and click Connect. 3. Log into your Solcast account at solcast.com and go to API Toolkit to generate a new API key. 4. Copy the key and paste it into the Switchy connection dialog, then click Authorize. 5. Return to any Space and type '@Solcast get resources' to confirm the connection works — you should see a list of your configured PV sites or an empty array if you haven't created any yet. 6. To invoke a tool, @mention Solcast followed by your request: '@Solcast get historic radiation for latitude -33.86 longitude 151.21 from 2024-01-01 to 2024-01-15' will return solar data for that Sydney location and date range.
What teammates see: by default, memories from Solcast are scoped to the Space (PROJECT visibility) - you can mark any memory PRIVATE or share it ORG-wide.
Works well with
Top models
Compatibility data appears once enough Spaces have used this MCP together with a given model.
How Switchy teams use it
Starter prompts
Historical Solar Data
@Solcast get historic radiation and weather for latitude 37.7749 longitude -122.4194 from 2024-03-01 to 2024-03-14Open in a Space →
List Configured Sites
@Solcast list all PV power resources available to my accountOpen in a Space →
Create New PV Resource
@Solcast create a new PV power resource at latitude 34.0522 longitude -118.2437 with 500 kW capacity, 25 degree tilt, 180 degree azimuth, and 0.85 system loss factorOpen in a Space →
Fetch Resource Details
@Solcast get the full configuration for PV power resource ID abc123-def456Open in a Space →
Update Panel Orientation
@Solcast update PV power resource ID abc123-def456 to change tilt to 30 degrees and azimuth to 200 degreesOpen in a Space →
Example outputs
Illustrative - representative of the model's voice and quality, not literal recordings.
@Solcast what was the solar irradiance and temperature at our Brisbane site (lat -27.47, lon 153.03) for the last week?
Retrieved 7 days of historical data for Brisbane location. Average global horizontal irradiance was 5.2 kWh/m²/day, peaking at 6.8 kWh/m²/day on the clearest day. Ambient temperatures ranged from 18°C to 31°C. Cloud cover averaged 42%, with three days showing minimal cloud interference. This data covers the period from January 8-14, 2025, using Solcast's satellite-derived estimates with 1-2km spatial resolution.
This example demonstrates read-only historical weather retrieval, useful for post-installation analysis or model validation. Historical data is available from 2007 to 7 days ago, with a 31-day maximum query window. Requires API key authentication; free tier accounts have daily request limits that may restrict frequent polling.
@Solcast create a new PV resource for our 500kW rooftop array in Phoenix: 33.45°N, 112.07°W, south-facing at 20° tilt, with 0.85 system efficiency
Created PV power resource 'Phoenix-Rooftop-001' with ID pv_8x3k9m2n. Configuration registered: 500kW DC capacity, azimuth 180° (south), tilt 20°, system derating 0.85 to account for inverter losses and soiling. Resource is now active for Advanced PV Power forecasting. You can request power output forecasts for this site using the resource ID, which will incorporate Solcast's irradiance models with your specific array geometry and efficiency parameters.
This write operation creates a persistent resource in your Solcast account that enables site-specific power forecasts. The resource stores technical specs (orientation, losses, tracking type) so future forecast requests don't need to repeat them. Advanced PV Power entitlement required; resource count may be limited by subscription tier.
@Solcast compare the forecasted vs actual performance for all our registered PV sites over the past 3 days and identify which sites are underperforming
Analyzed 4 registered PV resources against historical actuals. Phoenix-Rooftop-001 performed 7% below forecast (likely soiling accumulation—last cleaning was 6 weeks ago). Denver-Carport-002 matched forecast within 2% error. Austin-Ground-003 exceeded forecast by 4% (cooler than modeled temperatures improved efficiency). Portland-Roof-004 underperformed by 12%—historical irradiance shows persistent cloud cover not captured in original forecast horizon. Recommend inspecting Phoenix and Portland sites for maintenance needs or shading issues.
This synthesis example combines resource listing, historical data retrieval, and AI reasoning to surface operational insights. It showcases how the MCP enables cross-site analysis without manual data export. Accuracy depends on forecast horizon (degradation beyond 3-5 days) and whether resource configurations match actual installation specs.
Use-case deep-dives
When Solcast MCP fits multi-site solar ops teams
A 6-person solar installation company managing 40+ residential and commercial sites needs daily production forecasts to schedule maintenance crews and warranty claims. The Solcast MCP is the right call here because the team can programmatically create and update PV power resources as new sites come online, then pull historic radiation data to validate underperforming arrays against expected output. The 6-tool scope covers the full CRUD cycle for site modeling without manual API wrangling. This breaks down if your team is under 10 sites—at that scale, Solcast's web dashboard is faster than teaching everyone the MCP workflow. If you're triaging 30+ sites weekly and already use Switchy for ops handoffs, the MCP saves 2-3 hours per week versus copy-pasting CSV exports.
Why procurement teams skip this MCP
A utility procurement analyst buying solar power from 15 independent producers needs to verify vendor production claims before signing PPAs. The Solcast MCP can retrieve historic radiation actuals and compare them to vendor forecasts, but the 31-day history window per query makes this a poor fit for multi-year contract validation. If you're checking a single site's last month against a vendor's promise, the MCP works fine. If you're auditing 5 years of data across a portfolio, you'll hit the API rate limit and spend more time batching requests than analyzing results. For one-off spot checks during contract negotiation, the MCP is faster than waiting for Solcast support to export a custom dataset. For systematic due diligence, pay for Solcast's bulk data service instead.
When academic teams outgrow the MCP toolset
A 3-person university research group studying solar variability across 200 global coordinates needs 15 years of irradiance data to model climate impact on PV efficiency. The Solcast MCP's historic radiation tool only reaches back to 2007 and caps each request at 31 days, which means 60+ API calls per location to assemble a full time series. The MCP is useful for exploratory work—grabbing a few months of data to test a hypothesis or validate a model's boundary conditions—but it's not built for bulk dataset assembly. If your research question needs fewer than 10 locations and less than a year of history, the MCP keeps you in Switchy's shared workspace instead of switching contexts to Python notebooks. Beyond that threshold, script the Solcast API directly or use their academic data partnership program.
Frequently asked
What does the Solcast MCP do in Switchy?
It lets your team query solar irradiance forecasts and historical weather data, plus manage PV power resource configurations. You can pull radiation estimates for any location back to 2007, create new solar site models with panel specs, and update capacity or orientation parameters. Useful for energy modeling, site planning, or integrating solar data into your workflows without writing API code.
Do I need a paid Solcast account to use this MCP?
Yes. You need an active Solcast subscription and an API key from your account dashboard. Paste that key into Switchy's connection flow. The MCP respects your plan's rate limits and data entitlements, so if you're on a free tier you'll see restricted access to certain forecast horizons or resource counts.
Can the Solcast MCP generate real-time power output forecasts?
It retrieves forecast and historical irradiance data, but it doesn't run live power simulations inside Switchy. You create or update PV resource definitions with panel specs, then fetch the radiation estimates. Your team still needs to apply those numbers to your own power models or dashboards. Think of it as the data feed, not the calculator.
How is this different from calling Solcast's API directly?
The MCP wraps Solcast's REST endpoints so you can query forecasts or manage resources in plain English, without writing curl commands or managing tokens in code. You lose some flexibility—no custom HTTP headers or batch scripting—but you gain speed for ad-hoc questions and easier onboarding for non-developers on your team.
Who on the team should connect the Solcast MCP?
Whoever owns your Solcast API key. That's usually an engineer or data lead with access to your organization's Solcast dashboard. Once connected in Switchy, any team member can ask for forecasts or site data without needing their own key. Just make sure your Solcast plan allows the query volume your team will generate.