Struct NewDerivedInsight 

pub struct NewDerivedInsight {
    pub collective_id: CollectiveId,
    pub content: String,
    pub embedding: Option<Vec<f32>>,
    pub source_experience_ids: Vec<ExperienceId>,
    pub insight_type: InsightType,
    pub confidence: f32,
    pub domain: Vec<String>,
}

Input for creating a new derived insight.

The embedding field is required when using the External embedding provider, and optional when using the Builtin provider (which generates embeddings from content automatically).

Example

use pulsedb::{NewDerivedInsight, InsightType};

let insight = NewDerivedInsight {
    collective_id,
    content: "Error handling patterns converge on early return".to_string(),
    embedding: Some(embedding_vec),
    source_experience_ids: vec![exp_a, exp_b, exp_c],
    insight_type: InsightType::Pattern,
    confidence: 0.85,
    domain: vec!["rust".to_string(), "error-handling".to_string()],
};
let id = db.store_insight(insight)?;

Fields

collective_id: CollectiveId

The collective to store this insight in.

content: String

The insight content (text, max 50KB).

embedding: Option<Vec<f32>>

Pre-computed embedding vector (required for External provider).

source_experience_ids: Vec<ExperienceId>

IDs of the source experiences this insight was derived from (1-100).

insight_type: InsightType

The type of derivation.

confidence: f32

Confidence in this insight (0.0-1.0).

domain: Vec<String>

Domain tags for categorical filtering.