This quickstart shows you how to answer queries and search and retrieve developer documentation with the Developer Knowledge API by using the Google Cloud CLI. ## Before you begin Before you start using the Developer Knowledge API with the gcloud CLI, complete the following steps. ### Install and configure the gcloud CLI To install and configure the gcloud CLI, complete the following steps: 1. If you haven't installed the gcloud CLI, then [install the gcloud CLI](https://docs.cloud.google.com/sdk/docs/install). 2. Run the [`gcloud components update` command](https://docs.cloud.google.com/sdk/gcloud/reference/components/update) to make sure that you have the latest version: ``` gcloud components update ``` 3. Sign in to your Google Cloud account by running the [`gcloud auth login` command](https://docs.cloud.google.com/sdk/gcloud/reference/auth/login): ``` gcloud auth login ``` 4. Set your active Google Cloud project by running the [`gcloud config set` command](https://docs.cloud.google.com/sdk/gcloud/reference/config/set): ``` gcloud config set project PROJECT_ID ``` Replace `PROJECT_ID` with the ID of your Google Cloud project. ### Enable the API To enable the Developer Knowledge API, complete the following steps: 1. Enable the Developer Knowledge API in your Google Cloud project by running the [`gcloud services enable` command](https://docs.cloud.google.com/sdk/gcloud/reference/services/enable): ``` gcloud services enable developerknowledge.googleapis.com ``` You don't need specific Identity and Access Management (IAM) roles to enable or use the API. 2. Verify that the API is enabled for your project by running the [`gcloud services list` command](https://docs.cloud.google.com/sdk/gcloud/reference/services/list): ``` gcloud services list --enabled \ --filter="name:developerknowledge.googleapis.com" ``` The output lists the API name and title: ``` NAME TITLE developerknowledge.googleapis.com Developer Knowledge API ``` ## Generate answers from documentation The [`gcloud developer-knowledge answer-query` command](https://docs.cloud.google.com/sdk/gcloud/reference/developer-knowledge/answer-query) lets you ask questions about Google products and receive direct, natural-language answers. The command pulls information from official documentation sources and provides citations to the referenced pages. To ask a question and generate an answer, complete the following steps: 1. Run the following command to ask how to create a Cloud Storage bucket: ``` gcloud developer-knowledge answer-query \ --query="How do I create a Cloud Storage bucket?" ``` 2. Confirm that the command returns a generated answer and source references in YAML format: ```yaml answer: answerText: |- To create a Cloud Storage bucket, you can use the Google Cloud console, the gcloud CLI (`gcloud storage buckets create`), client libraries, or the REST API... citations: - endIndex: 158 sources: - referenceIndex: 0 startIndex: 0 references: - documentReference: documentChunk: content: |- This document shows you how to create a Cloud Storage [bucket](https://docs.cloud.google.com/storage/docs/buckets)... document: dataSource: docs.cloud.google.com name: documents/docs.cloud.google.com/storage/docs/creating-buckets title: Create a bucket uri: https://docs.cloud.google.com/storage/docs/creating-buckets parent: documents/docs.cloud.google.com/storage/docs/creating-buckets ``` The `answer` object in the output includes the following fields: - `answerText`: the generated natural-language answer to your query. - `citations`: byte-offset ranges, `startIndex` and `endIndex`, in `answerText` that map to supporting entries in `references` by `referenceIndex`. - `references`: the source documentation chunks and metadata, including `document` and `parent`, used to generate the answer. ## Search for document chunks To find specific text excerpts in Google's developer documentation rather than a generated answer, use the [`gcloud developer-knowledge documents search-chunks` command](https://docs.cloud.google.com/sdk/gcloud/reference/developer-knowledge/documents/search-chunks). This command scans the documentation corpus and returns matching content chunks alongside the resource names of their parent documents. To search for document chunks, complete the following steps: 1. Run the following command to search for documentation about creating Cloud Storage buckets: ``` gcloud developer-knowledge documents search-chunks \ --query="How do I create a Cloud Storage bucket?" ``` 2. Confirm that the command returns a list of matching document chunks in YAML format: ```yaml --- content: |- This document shows you how to create a Cloud Storage [bucket](https://docs.cloud.google.com/storage/docs/buckets). If not otherwise specified in your request, buckets are created in the [US multi-region](https://docs.cloud.google.com/storage/docs/locations)... document: contentLengthBytes: 31842 dataSource: docs.cloud.google.com description: World-wide storage and retrieval of data in Google Cloud. name: documents/docs.cloud.google.com/storage/docs/creating-buckets title: Create a bucket updateTime: '2026-09-10T20:05:41Z' uri: https://docs.cloud.google.com/storage/docs/creating-buckets view: DOCUMENT_VIEW_BASIC id: c1 parent: documents/docs.cloud.google.com/storage/docs/creating-buckets relevanceScore: 0.856608 ``` Each chunk in the output includes the following fields: - `content`: the matched text snippet from the documentation. - `document`: metadata about the source document, including its `title`, `description`, `uri`, `dataSource`, and `updateTime`. - `id`: the identifier of the chunk within the document. - `parent`: the resource name of the parent document. You can pass this value to the [`describe` command](https://docs.cloud.google.com/sdk/gcloud/reference/developer-knowledge/documents/describe) to retrieve the full document. - `relevanceScore`: the relevance score of the chunk to the search query. ## Retrieve a document After you find a relevant document chunk, use the `parent` field from those search results to fetch the complete Markdown content of the document. Run the [`gcloud developer-knowledge documents describe` command](https://docs.cloud.google.com/sdk/gcloud/reference/developer-knowledge/documents/describe) with the resource name of the document. For example, to retrieve the document on creating Cloud Storage buckets, complete the following steps: 1. Run the following command: ``` gcloud developer-knowledge documents describe \ documents/docs.cloud.google.com/storage/docs/creating-buckets ``` 2. Confirm that the command returns the metadata and full Markdown content of the document: ```yaml content: | This document shows you how to create a Cloud Storage [bucket](https://docs.cloud.google.com/storage/docs/buckets). If not otherwise specified in your request, buckets are created in the [`US` multi-region](https://docs.cloud.google.com/storage/docs/locations) with a default storage class of [Standard storage](https://docs.cloud.google.com/storage/docs/storage-classes) and have a seven-day [soft delete](https://docs.cloud.google.com/storage/docs/soft-delete) retention duration... contentLengthBytes: 31842 dataSource: docs.cloud.google.com description: World-wide storage and retrieval of data in Google Cloud. name: documents/docs.cloud.google.com/storage/docs/creating-buckets title: Create a bucket updateTime: '2026-09-10T20:05:41Z' uri: https://docs.cloud.google.com/storage/docs/creating-buckets view: DOCUMENT_VIEW_CONTENT ``` The output includes the following fields: - `content`: the full Markdown text of the document. - `contentLengthBytes`: the total size of the document content in bytes. - `dataSource`: the documentation domain hosting the document. - `description`: a brief summary of the document. - `name`: the unique resource name of the document. - `title`: the title of the document. - `updateTime`: the timestamp when the document was last updated. - `uri`: the public URL of the documentation page. - `view`: the document view returned (`DOCUMENT_VIEW_CONTENT`, `DOCUMENT_VIEW_BASIC`, or `DOCUMENT_VIEW_FULL`). ## What's next - Learn how to [generate answers from documentation](https://developers.google.com/knowledge/answer-query.md.txt). - Learn how to [filter search results, paginate, and specify document views](https://developers.google.com/knowledge/search-retrieve). - Explore the [corpus reference](https://developers.google.com/knowledge/reference/corpus-reference.md.txt) to view the full list of included documentation. - Refer to the [`gcloud developer-knowledge` reference](https://docs.cloud.google.com/sdk/gcloud/reference/developer-knowledge) for details on all available commands and flags. - Review the [REST API reference](https://developers.google.com/knowledge/reference/rest.md.txt) for complete specifications for API methods. - Learn how to [set up the Model Context Protocol (MCP) server in Google Antigravity](https://codelabs.developers.google.com/developer-knowledge-mcp-antigravity).