Install one server — uvx solar-data-mcp — and your agent gets all 18 tools, 11 skills that teach it how to use them, and 4 ready-made report prompts, across four data domains: production modeling, economics, market data, and forecasts. Every tool returns the same envelope — data + units + source + assumptions + warnings — so the agent always knows what a number means, where it came from, and which defaults were injected on its behalf.

What can you ask it?

• Thinking about solar at home — "Would a 6 kW system pay off at my house?" · "Is this $21,000 quote fair?" · "What will my array generate tomorrow?"
• Selling or installing solar — "Build a proposal for this customer" · "How long does permitting take in Phoenix?" · "Where should we expand next?"
• Studying the market — "Brief me on the Texas solar market" · "How have installed prices trended in Colorado?" · "Which big plants have batteries?"

Behind each question the agent picks the right tools (or a skill routes it), and every number comes back with units, a source, and the assumptions made on your behalf.

Quickstart

1. Get a free NREL API key — https://developer.nlr.gov/signup/ (or use DEMO_KEY to try it out — 10 requests/hour).

2. Add the server to your agent — snippets for every major agent below; for Claude Desktop, merge this into claude_desktop_config.json (full example in examples/):

   config.json

   Copy

   {
     "mcpServers": {
       "solar-data": {
         "command": "uvx",
         "args": ["solar-data-mcp"],
         "env": {
           "NREL_API_KEY": "YOUR_KEY_HERE",
           "OPENEI_API_KEY": "YOUR_KEY_HERE",
           "EIA_API_KEY": "YOUR_KEY_HERE"
         }
       }
     }
   }
   

   Only NREL_API_KEY is needed to start: the server runs with any subset of keys, and a tool missing its key returns setup instructions instead of failing silently.

3. Restart your client and ask — try "Compare annual production for an 8 kW system in Mesa, AZ at 10° vs 25° tilt."

Verify keys and connectivity anytime:

$ uvx solar-data-mcp doctor

Add it to your agent

Claude Code

$ claude mcp add solar-data \
    --env NREL_API_KEY=YOUR_KEY --env OPENEI_API_KEY=YOUR_KEY --env EIA_API_KEY=YOUR_KEY \
    -- uvx solar-data-mcp

(or commit the quickstart JSON to your project's .mcp.json)

Codex CLI — ~/.codex/config.toml:

[mcp_servers.solar-data]
command = "uvx"
args = ["solar-data-mcp"]

[mcp_servers.solar-data.env]
NREL_API_KEY = "YOUR_KEY_HERE"
OPENEI_API_KEY = "YOUR_KEY_HERE"
EIA_API_KEY = "YOUR_KEY_HERE"

(or codex mcp add solar-data --env NREL_API_KEY=YOUR_KEY -- uvx solar-data-mcp; use env_vars = ["NREL_API_KEY"] to forward keys from your shell instead of hardcoding them)

OpenCode — opencode.json in your project (or ~/.config/opencode/opencode.json):

{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "solar-data": {
      "type": "local",
      "command": ["uvx", "solar-data-mcp"],
      "enabled": true,
      "environment": {
        "NREL_API_KEY": "YOUR_KEY_HERE",
        "OPENEI_API_KEY": "YOUR_KEY_HERE",
        "EIA_API_KEY": "YOUR_KEY_HERE"
      }
    }
  }
}

Hermes — ~/.hermes/config.yaml, then /reload-mcp:

mcp_servers:
  solar-data:
    command: "uvx"
    args: ["solar-data-mcp"]
    env:
      NREL_API_KEY: "YOUR_KEY_HERE"
      OPENEI_API_KEY: "YOUR_KEY_HERE"
      EIA_API_KEY: "YOUR_KEY_HERE"

Claude Desktop — the quickstart JSON above.

Anything else that speaks MCP (stdio) — command uvx, args ["solar-data-mcp"], keys in the env block.

API keys

All free, and every one optional — the server starts with none set.

Market tools (USPVDB, Tracking the Sun, SolarTRACE) and forecasts need no key. Full forecast output additionally needs the Quartz model installed into a persistent environment (see packages/solar-forecast/); without it the forecast tools return install instructions.

Servers

uvx solar-data-mcp — the install above — serves all four domains on one stdio entry. Each domain also ships as a standalone server:

Per-server config: examples/claude_desktop_config.per-server.json.

⚠️ Run the combined solar-data server or the per-domain servers, not both — and never solar-economics and solar-market side by side. Both open the same local DuckDB bulk store, which allows only one process at a time.

Tools

Eighteen tools across four domains; parameter details live in each tool's docstring and each domain package's README.

Skills & reports

Skills are procedures shipped inside the combined server that teach an agent to chain the tools correctly — ordering, sync prerequisites, honest reporting. They're MCP resources: skill://solar/index routes by question shape, skill://solar/<name> is the procedure. Grouped by who's asking:

• Homeowners — site assessment, quote review, performance check
• Installers — proposal builder, territory expansion
• Analysts — market brief, pricing analysis, utility-scale scout, incentive scan
• Cross-cutting — data sync (bulk snapshots), data conventions (envelope literacy)

Four of these render reports with a fixed document shape and are also exposed as MCP prompts your host surfaces natively — market_brief, site_assessment, quote_review, proposal_builder (in Claude Code: /mcp__solar-data__market_brief).

Full catalog, routing design, and report templates: docs/skills.md.

Development

$ git clone https://github.com/hoodsy/solar-data-mcp && cd solar-data-mcp
$ uv sync                # install the workspace
$ uv run pytest          # fixture replay only, no network

Layout: packages/core (shared HTTP client, cache, envelope), one package per domain server, and packages/solar-data-mcp (the umbrella that mounts all four on one stdio entry). Smallest possible client: examples/example_client.py.

License

MIT. Per-source data licensing/attribution is exposed as MCP resources (source://<name>/license).