1
import chromadb
2
import google.generativeai as genai
3
import os
4
from ingest import process_file # Importing previous logic
5

6
genai.configure(api_key=os.environ["GEMINI_API_KEY"])
7

8
client = chromadb.PersistentClient(path="./vault_data")
9
collection = client.get_or_create_collection(name="campaign_memory")
10

11
def add_to_vault(chunks, meta):
12
    for i, chunk in enumerate(chunks):
13
        res = genai.embed_content(
14
            model="models/text-embedding-004",
15
            content=chunk,
16
            task_type="retrieval_document"
17
        )
18
        collection.add(
19
            ids=[f"id_{meta['id']}_{i}"],
20
            embeddings=[res['embedding']],
21
            documents=[chunk],
22
            metadatas=[meta]
23
        )
24

25
# --- Test the Alchemy ---
26
if __name__ == "__main__":
27
    # Ensure initialization.mdx exists in your directory
28
    chunks = process_file("initialization.mdx")
29
    add_to_vault(chunks, {"id": "lesson_1", "type": "tutorial"})
30
    print(f"Alchemy Success: {len(chunks)} Vectors Locked in Vault.")

1
import { GoogleGenerativeAI } from "@google/generative-ai";
2
import { ChromaClient } from "chromadb";
3
import { ingestFile } from "./ingest.js"; // Importing previous logic
4

5
const genAI = new GoogleGenerativeAI(process.env.GEMINI_API_KEY);
6
const chroma = new ChromaClient();
7

8
async function addToVault(chunks, metadata) {
9
    const collection = await chroma.getOrCreateCollection({ name: "campaign_memory" });
10
    const model = genAI.getGenerativeModel({ model: "text-embedding-004" });
11

12
    for (let i = 0; i < chunks.length; i++) {
13
        const res = await model.embedContent(chunks[i]);
14
        await collection.add({
15
            ids: [`${metadata.id}_${i}`],
16
            embeddings: [res.embedding.values],
17
            documents: [chunks[i]],
18
            metadatas: [metadata]
19
        });
20
    }
21
}
22

23
// --- Test the Alchemy ---
24
(async () => {
25
    const chunks = await ingestFile("initialization.mdx");
26
    await addToVault(chunks, { id: "lesson_1", type: "tutorial" });
27
    console.log(`Alchemy Success: ${chunks.length} Vectors Locked in JS Vault.`);
28
})();

1
import dev.langchain4j.data.embedding.Embedding;
2
import dev.langchain4j.model.googleai.GoogleAiEmbeddingModel;
3
import dev.langchain4j.store.embedding.chroma.ChromaEmbeddingStore;
4

5
public class VaultTest {
6
    public static void main(String[] args) {
7
        // Assume Ingestor class from 2.2 is available
8
        Ingestor ingestor = new Ingestor();
9
        List<String> chunks = ingestor.processFile("raw markdown content here...");
10

11
        GoogleAiEmbeddingModel model = GoogleAiEmbeddingModel.builder()
12
            .apiKey(System.getenv("GEMINI_API_KEY"))
13
            .modelName("text-embedding-004")
14
            .build();
15

16
        ChromaEmbeddingStore store = ChromaEmbeddingStore.builder()
17
            .baseUrl("http://localhost:8000")
18
            .collectionName("campaign_memory")
19
            .build();
20

21
        for (String chunk : chunks) {
22
            Embedding embedding = model.embed(chunk).content();
23
            store.add(embedding, chunk);
24
        }
25
        System.out.println("Alchemy Success: Java Vault Updated and Tested.");
26
    }
27
}