Predict Outcome

Predict outcomes given causes (Interventional reasoning).

This implements the second level of Pearl's Causal Ladder: "What if I DO X?" (vs. just observing)

Given a set of causes, predicts the most likely effect by:

1. Finding all causal paths from causes to target
2. Weighting paths by edge strength
3. Estimating probability of target outcome

Why This Matters

Planning requires interventional reasoning. We can:

• Ask "If I take action X, what will happen?"
• Trace causal chains to predict downstream effects
• Identify necessary conditions for goals

Args: request: Cause IDs and target to predict current_user: Authenticated user db: Database session

Returns: Predicted outcome with probability and causal chain

Example: ``` Causes: ["memory_user_said_hello", "memory_i_responded"] Target: "user_satisfaction"

Returns:
{
    "outcome_id": "user_satisfaction",
    "probability": 0.78,
    "confidence": 0.85,
    "causal_chain": ["said_hello", "responded", "user_satisfaction"]
}
```