npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2026 – Pkg Stats / Ryan Hefner

@thenajs/qdrant-client

v0.6.0

Published

Cliente Qdrant nativo para ThenaJS — implementação de VectorStore sobre a API REST, sem SDK.

Readme

@thenajs/qdrant-client

Cliente Qdrant nativo para ThenaJS — implementação de VectorStore sobre a API REST. Sem SDK: fetch puro, herdando o retry e o timeout do transporte do framework.

Requer Qdrant 1.10 ou superior. O endpoint unificado /points/query, que o cliente usa para buscar, estreou nessa versão.

Instalação

npm install @thenajs/qdrant-client

Uso

import { QdrantStore } from "@thenajs/qdrant-client";

export class MeuQdrant extends QdrantStore {
  constructor() {
    super({
      url: "http://localhost:6333",
      collection: "conhecimento",
      retry: { maxAttempts: 3, timeoutMs: 10_000 },
    });
  }
}

Registre uma vez no config:

export const config: ThenaConfig = {
  memory: [MeuQdrant],
};

E injete em qualquer agente:

@Agent({
  provider: LocalOllamaProvider,   // o embed() dele gera os vetores
  prompt: "./assistente.agent.md",
})
export class Assistente {
  constructor(private readonly memory: VectorMemory) {}

  async beforePrompt(prompt: string) {
    const achados = await this.memory.recall("como faço deploy?", {
      dataset: "persistent",
      limit: 3,
    });
    return `${prompt}\n\n${achados.map((a) => `- ${a.text}`).join("\n")}`;
  }
}

Documentação completa em thenajs.github.io.

Credentials

| Campo | Default | O que faz | | --- | --- | --- | | url | — | Endereço do Qdrant (obrigatório) | | apiKey | — | Header api-key, para Qdrant Cloud | | collection | "thena_memory" | A collection onde tudo é gravado | | datasetField | "dataset" | Campo do payload que particiona | | retry | ligado | Política de retry/timeout do HttpTransport |

Uma collection, vários contextos

O dataset não vira collection. É um campo do payload, com índice dedicado — que é a recomendação do próprio Qdrant: muitas collections geram overhead de recursos, e o Qdrant Cloud limita a 1000 por cluster.

Na prática isso significa que dá para buscar dentro de um contexto ou através de todos:

await this.memory.recall("pergunta", { dataset: "persistent" });  // um dataset
await this.memory.recall("pergunta");                             // o "default"
await this.memory.recall("pergunta", { dataset: null });          // todos

Em Qdrant 1.12+, o índice é criado com is_tenant: true, que co-loca os pontos do mesmo dataset em disco. Em versões anteriores o cliente cai para o índice keyword simples — funciona igual, só sem essa otimização.

Licença

MIT