JSPM – gatsby-plugin-recommend-article@0.1.6

Package Exports

gatsby-plugin-recommend-article
gatsby-plugin-recommend-article/index.js

This package does not declare an exports field, so the exports above have been automatically detected and optimized by JSPM instead. If any package subpath is missing, it is recommended to post an issue to the original package (gatsby-plugin-recommend-article) to support the "exports" field. If that is not possible, create a JSPM override to customize the exports field for this package.

Readme

Gatsby plugin to recommend articles based on OpenAI embeddings API and Qdrant vector search.

View the demo site on GitHub Pages
View the demo repository

Install

with npm

npm install gatsby-plugin-recommend-article

with yarn

yarn add gatsby-plugin-recommend-article

with pnpm

pnpm add gatsby-plugin-recommend-article

with bun

bun add gatsby-plugin-recommend-article

Quick start

Add the plugin to your gatsby-config.js

module.exports = {
  plugins: [
    {
      resolve: `gatsby-plugin-recommend-article`,
      options: {
        qdrant: {
          url: "<your-qdrant-url>",
        },
        openai: {
          apiKey: "<your-openai>",
        },
        limit: 3,
        toPayload: ({ node }) => {
          return {
            title: node.frontmatter.title,
            content: node.excerpt,
            tags: node.frontmatter.tags,
          };
        },
      },
    },
  ],
};

Query for recommended articles

query {
  allMarkdownRemark(filter: { id: { eq: "xxxx" } }) {
    nodes {
      id
      excerpt
      recommends {
        id
        excerpt
      }
    }
  }
}

{
  "data": {
    "allMarkdownRemark": {
      "nodes": [
        {
          "id": "xxxx",
          "excerpt": "...",
          "recommends": [
            {
              "id": "yyyy",
              "excerpt": "..."
            },
            {
              "id": "zzzz",
              "excerpt": "..."
            },
            ...
          ]
        },
      ]
    }
  }
}

Options

Name	Type	Description	Default	Required
qdrant	object	Configuration for Qdrant.	-	✅
openai	object	Configuration for OpenAI.	-	❌
limit	number	Maximum number of recommended articles.	5	❌
nodeType	string	Type of node to add `recommends` field. Also, the type of `recommends` is an array of the specified type.	"MarkdownRemark"	❌
toPayload	function	Function to convert node to payload. payload is only used to generate vector.	`(node) => JSON.stringify({ body: node.excerpt ?? "" })`	❌

qdrant

Name	Type	Description	Default	Required
url	string	URL of the Qdrant server	-	✅
apiKey	string	API key for authenticating with Qdrant	-	❌
https	boolean	Whether to use HTTPS	false	❌
headers	object	Additional headers for requests	`{}`	❌
onDisk	boolean	Whether to use on-disk storage	false	❌
collectionName	string	Name of the Qdrant collection to use	"articles"	❌

openai

Name	Type	Description	Default	Required
baseURL	string	base URL for the OpenAI API	-	❌
apiKey	string	API key for authenticating with OpenAI	-	❌
organization	string	ID of the OpenAI organization	-	❌
project	string	ID of the OpenAI project	-	❌
embeddingModel	string	Model used for generating embeddings	"text-embedding-3-small"	❌
embeddingSize	number	Size of the embedding vector	1536	❌

Contributing

Feel free to open a PR or an Issue.

However, you must promise to follow our Code of Conduct.

See here for more details on contributing.

License

gatsby-plugin-recommend-article released under the MIT License