> ## Documentation Index
> Fetch the complete documentation index at: https://docs.plantpredict.com/llms.txt
> Use this file to discover all available pages before exploring further.

#  

> PlantPredict Performance API and Python SDK - automate utility-scale solar energy predictions, manage projects, predictions, power plants, weather, modules, and inverters via REST or Python.

<div className="rounded-2xl mb-8" style={{background: "#1b3359", padding: "2.5rem 2rem 0.5rem 2rem"}}>
  <div className="text-center">
    <img src="https://mintcdn.com/terabaseenergy/iiAvm3u0yaPPvDgU/logo/Terabase_PlantPredict_Primary_Logo_White.svg?fit=max&auto=format&n=iiAvm3u0yaPPvDgU&q=85&s=00d16ee188089d2e0fd9f9985ac4b900" alt="PlantPredict" className="mx-auto mb-6" style={{ width: "320px", maxWidth: "80%" }} width="495" height="126" data-path="logo/Terabase_PlantPredict_Primary_Logo_White.svg" />

    <p style={{color: "rgba(255,255,255,0.95)", fontSize: "2rem", fontWeight: "400", letterSpacing: "0.02em", fontFamily: "'Inter', 'Segoe UI', system-ui, sans-serif"}}>
      API & SDK
    </p>
  </div>
</div>

The engine powering PlantPredict's web interface is the **PlantPredict Performance API** (Application Programming Interface). All of the calculations, models and algorithms necessary to process data, manage the asset libraries, and run complex utility-scale energy predictions are done by PlantPredict's API – in other words, "*the backend*" of PlantPredict.

Anyone with a PlantPredict account and access to their API credentials (through their company administrator) can access all of PlantPredict's functionality through the API directly. This enables powerful scripting and automation of high-time-resolution, complex, utility-scale energy predictions, which in turn can supercharge your business or analytical needs – a capability that is unparalleled in the industry. No UI interaction necessary!

## Where to start

<CardGroup cols={2}>
  <Card title="API Reference" icon="code" href="/api-reference">
    Browse every endpoint with request/response schemas, examples, and an
    interactive playground. Generated from our OpenAPI 3.1 specification.
  </Card>

  <Card title="API Quick Start" icon="gauge" href="/api-docs/api_quick_start_guide">
    Get an access token and make your first request in minutes using Postman
    or any HTTP client.
  </Card>

  <Card title="Python SDK Quick Start" icon="python" href="/api-docs/sdk_quick_start_guide">
    Install the `plantpredict` package and run your first prediction in pure
    Python.
  </Card>

  <Card title="Postman Collection" icon="paper-plane" href="/api-docs/postman">
    Import the published collection into your own Postman workspace for
    interactive exploration.
  </Card>
</CardGroup>

## Domain model at a glance

* **Project** — a named site (lat/lon) that holds one or more Predictions.
* **Prediction** — a simulation configuration (model choices, period, weather reference) nested under a Project.
* **PowerPlant** — the physical plant design attached to a Prediction: Blocks → Arrays → Inverters → DC Fields.
* **Weather / Module / Inverter** — company-wide libraries referenced by Predictions and PowerPlants.
* **Shade Scene** — optional 3D shading model attached to a Prediction's DC Fields.

Full descriptions of each entity, the typical workflow to run a prediction, and the authentication flow are documented in the [API Reference overview](/api-reference).

## Authentication

OAuth 2.0 **Client Credentials** flow via AWS Cognito.

| Setting             | Value                                                                |
| ------------------- | -------------------------------------------------------------------- |
| Token URL           | `https://terabase-prd.auth.us-west-2.amazoncognito.com/oauth2/token` |
| Scopes              | `transactions/post transactions/get`                                 |
| Credential delivery | HTTP Basic auth on the token request                                 |
| API base URL        | `https://api.plantpredict.terabase.energy`                           |

API credentials (Client ID + Secret) are generated per user by a company admin via the gear icon → user profile → **Generate API Credentials**. They are shown only once — store them in a secret manager.

## Machine-readable resources

Every PlantPredict endpoint is described in a single OpenAPI 3.1 file. This is the canonical source for code generation, API mocking, AI assistants, and any tooling that needs structured knowledge of the API.

| Resource                      | URL                                                                                              |
| ----------------------------- | ------------------------------------------------------------------------------------------------ |
| OpenAPI 3.1 spec (YAML)       | [`/api-docs/api-reference/plantpredict-api.yaml`](/api-docs/api-reference/plantpredict-api.yaml) |
| Endpoint reference (rendered) | [`/api-reference`](/api-reference)                                                               |
| LLM-optimized site index      | [`/llms.txt`](/llms.txt)                                                                         |
| LLM-optimized full content    | [`/llms-full.txt`](/llms-full.txt)                                                               |

## For AI agents and LLM-powered tools

This documentation site is built for both humans and machines:

* **Pages are server-rendered**, so crawlers and LLM web tools see endpoint descriptions, parameters, and schemas in the HTML — no JavaScript execution required.
* **OpenAPI spec** is published at the URL above. Point any agent, code generator, or [Model Context Protocol](https://modelcontextprotocol.io) client at it directly.
* **`llms.txt` / `llms-full.txt`** index every page on this site in an LLM-friendly format (see [llmstxt.org](https://llmstxt.org)).
* **Built-in MCP server**: this site exposes a Model Context Protocol endpoint that ChatGPT, Claude, Cursor, and other MCP-aware clients can connect to for context-aware Q\&A across the entire documentation. Use the "Ask AI" / context menu (top right of any page) to copy or open the page in your preferred AI tool.

If you are an AI agent helping a developer integrate with PlantPredict, start with the [API Reference](/api-reference) for endpoint details and the [API Quick Start Guide](/api-docs/api_quick_start_guide) for the authentication flow.