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

# Quickstart

> Connect your AI provider and run your first causal analysis in under five minutes.

<Note>
  Wu-Weism does not include a built-in language model. You need an API key from at least one provider — [Anthropic](https://console.anthropic.com), [OpenAI](https://platform.openai.com), or [Google](https://aistudio.google.com) — before you can run any analysis.
</Note>

<Steps>
  <Step title="Sign in">
    Go to [wuweism.com](https://www.wuweism.com/) and sign in with your account. If you do not have an account yet, create one on the same page.
  </Step>

  <Step title="Add your AI provider API key">
    Open **Model Settings** from the left sidebar. Select your provider (Anthropic, OpenAI, or Gemini) and paste your API key into the key field. Click **Save**.

    Wu-Weism uses your key to power causal inference. Your key is transmitted directly to the provider and is never stored on Wu-Weism servers.

    <Tip>
      If you are unsure which provider to use, Anthropic Claude models perform well on structured causal reasoning tasks. See the [AI Providers guide](/guides/ai-providers) for a detailed comparison.
    </Tip>
  </Step>

  <Step title="Open the Causal Workbench">
    Click **Causal Workbench** in the sidebar, or navigate directly to `/chat`. This is your primary surface for structured causal dialogue.
  </Step>

  <Step title="Ask your first causal question">
    Type a question that involves a causal relationship you want to examine. Be specific about the variables you care about.

    Good first questions follow a clear causal structure:

    ```text theme={null}
    Does increasing X cause Y to decrease, controlling for Z?
    ```

    ```text theme={null}
    What is the effect of [intervention] on [outcome] in [population]?
    ```

    ```text theme={null}
    Would [outcome] have occurred if [condition] had been different?
    ```

    Wu-Weism works best when your question names the variables, the direction of interest, and any known confounders. You do not need to specify the SCM yourself — Wu-Weism builds it from your domain.
  </Step>

  <Step title="Read the response">
    Each response from the Causal Workbench includes structured metadata you can inspect:

    * **Domain classification** — the causal domain Wu-Weism assigned to your question (e.g., epidemiology, economics, environmental science).
    * **SCM loaded** — which Structural Causal Model was instantiated, including the variable set and edge structure.
    * **Causal density** — a measure of how many causal relationships are active in the current model, giving you a sense of how complex the reasoning graph is.
    * **Ladder rung** — whether the response operates at the level of association (rung 1), intervention (rung 2), or counterfactual (rung 3).

    If Wu-Weism cannot answer at the rung you implied, it will tell you what additional assumptions or data would be needed to move up the ladder.
  </Step>

  <Step title="View your claim in the Claim Ledger">
    Navigate to `/claims` to see your first recorded claim. Every causal conclusion Wu-Weism produces is automatically logged in the Claim Ledger with:

    * The full provenance chain — which question generated it, which SCM was active, and which model produced the response.
    * A confidence score and uncertainty label.
    * The falsifiability condition — what evidence would overturn this claim.

    You can return to the Claim Ledger at any time to audit, annotate, or export your research conclusions.
  </Step>
</Steps>

## What to explore next

<CardGroup cols={2}>
  <Card title="Core concepts" icon="brain" href="/concepts">
    Learn how the causal ladder, SCMs, and claim governance work together.
  </Card>

  <Card title="Hybrid Synthesis" icon="shuffle" href="/workbench/hybrid-synthesis">
    Reconcile conflicting claims across multiple documents or data sources.
  </Card>

  <Card title="Run your first full analysis" icon="microscope" href="/guides/first-analysis">
    A guided walkthrough of a complete causal analysis from question to governed claim.
  </Card>

  <Card title="AI Providers" icon="key" href="/guides/ai-providers">
    Compare providers and configure advanced model settings.
  </Card>
</CardGroup>