The easy-to-use open source Business Intelligence and Embedded Analytics tool that lets everyone work with data :bar_chart:
TypeScript analyticsbibusiness-intelligencebusinessintelligenceclojuredashboarddatadata-analysisdata-visualizationdatabasemetabasemysqlpostgrespostgresqlreportingslacksql-editorvisualization
docs - CLI page and revisions on agent-driven workflow (#74976)
Jeff Bruemmerb9113d674151
diff --git a/docs/README.md b/docs/README.mdindex 506ef5040ef..8e94eab8dc8 100644--- a/docs/README.md+++ b/docs/README.md@@ -52,7 +52,8 @@ Metabase's reference documentation.- [Development instances](./installation-and-operation/development-instance.md)- [Serialization](./installation-and-operation/serialization.md)- [Remote sync](./installation-and-operation/remote-sync.md)-- [Commands](./installation-and-operation/commands.md)+- [Metabase CLI](./installation-and-operation/metabase-cli.md)+- [Metabase JAR commands](./installation-and-operation/commands.md)- [Supported browsers](./installation-and-operation/supported-browsers.md)- [Privacy](./installation-and-operation/privacy.md)- [About the anonymous usage data we collect](./installation-and-operation/information-collection.md)diff --git a/docs/ai/file-based-development.md b/docs/ai/file-based-development.mdindex 179ea0d6626..2dad6381370 100644--- a/docs/ai/file-based-development.md+++ b/docs/ai/file-based-development.md@@ -1,33 +1,35 @@---title: Agent-driven development-summary: "Use an agent with skills to create Metabase content on your local machine as YAML files, then upload the files to your Metabase."+summary: "Use a coding agent and the Metabase CLI to create Metabase content, then version that content as YAML files with Remote Sync."---# Agent-driven development{% include plans-blockquote.html feature="Agent-driven development" %}-Metabase content like questions and dashboards can be serialized as YAML files. You can edit those YAML files by hand, sure, but now that we have actual genies, you can just ask them to edit them for you (call it "lamp-rubbing development").+Now that we have actual genies, you can just ask the genies to create the content for you (call it "lamp-rubbing development").-Paired with some skills we've developed, you can export your Metabase as YAML files, ask your agent to create new questions and dashboards YAML files, then import that new content into your Metabase. It's a whole new way to work with Metabase.+Give an agent the [Metabase CLI](../installation-and-operation/metabase-cli.md), and it can create content for you. But since LLMs can be unpredictable, we recommend developing content in a dev instance of Metabase, then using [Remote Sync](../installation-and-operation/remote-sync.md) to get your changes into production.-## The agent-driven development toolkit--To develop your Metabase content on your local filesystem, we've put together a set of tools, including a set of agent skills.--- [**Metabase Representation Format**](https://github.com/metabase/representations): the YAML schema and spec for every Metabase entity (questions, dashboards, collections, transforms, and so on).-- **[Metabase Database Metadata Format](https://github.com/metabase/database-metadata)**: diff-friendly representations of synced databases, their tables, and their fields, as a tree of YAML files.-- **Export and Import** CLI and API endpoints to move serialized content between your local files and Metabase.-- [**Remote Sync**](../installation-and-operation/remote-sync.md) (Optional): push content from a Read-write Metabase into a git repo, and pull it into a Read-only Metabase in production.+With this set up, a typical workflow using an agent with a development instance of Metabase would be:-## How content moves between files and Metabase+1. Prompt the agent with `/metabase-cli Create a dashboard based on the sales table.`+2. Agent creates questions and a dashboard.+3. View the dashboard in your dev instance.+4. Iterate either in your Metabase or via the agent until you're happy with the dashboard.+5. Use Remote Sync to push your changes to a repo.+6. Create a PR.+7. Merge the changes.+8. Once merged, your production Metabase pulls in the changes via Remote Sync.-You'll need a way to get YAML files out of Metabase to edit and back into Metabase to verify and ship. There are two options:+## The agent-driven development toolkit-- **[Remote Sync](../installation-and-operation/remote-sync.md)** — push and pull from inside Metabase. Requires a Read-write development instance and a Read-only production instance.-- **Serialization API** — `curl`-based export and import against the `/api/ee/serialization/` endpoints.+To develop your Metabase content with an agent, we've put together a set of tools.-Pick one before you start the [Initial setup](#initial-setup); the setup steps differ slightly (Remote Sync doesn't need a separate API key in production).+- A Metabase instance to use for development.+- [**Metabase CLI**](../installation-and-operation/metabase-cli.md): a command-line client (`mb`) your agent uses to create content directly in your Metabase. Use the CLI with the [`/metabase-cli` skill](https://github.com/metabase/agent-skills/tree/main/skills/metabase-cli).+- [**Metabase Representation Format**](https://github.com/metabase/representations): the YAML schema and spec for every Metabase entity (questions, dashboards, collections, transforms, and so on). This is the format your content takes once you version it as files.+- [**Remote Sync**](../installation-and-operation/remote-sync.md): push content from a Read-write Metabase into a git repo, and pull it into a Read-only Metabase in production.## Initial setup@@ -35,20 +37,19 @@ Some things to put into place to get a workflow going:1. [Set up a development Metabase](#set-up-a-development-metabase)2. [Set up a repository to version your YAML files](#set-up-a-repository-to-version-your-yaml-files)-3. [Add agent skills to your repository](#add-agent-skills-to-your-repository)-4. [Download the database metadata](#download-the-database-metadata)+3. [Install and authenticate the Metabase CLI](#install-and-authenticate-the-metabase-cli)+4. [Add the agent skill](#add-the-agent-skill)-Once you have these set up, you can step through one of the example workflows.+Once you have these set up, you can step through the example workflow.### Set up a development Metabase1. Set up a Metabase instance to check your work before pushing changes to production. This Metabase should connect to the same data warehouse(s) your production Metabase connects to. A [config file](../configuring-metabase/config-file.md) will come in handy here.-2. Create an [API key](../people-and-groups/api-keys.md#create-an-api-key) and assign it to the Admin group. The skills export and import all content and read database metadata, so they need Admin-level access. If you're using the [Serialization API workflow](#how-content-moves-between-files-and-metabase), you'll also need to create an API key in your production Metabase so you can import your files into it.+2. Create an [API key](../people-and-groups/api-keys.md#create-an-api-key) in this development Metabase and assign it to the Admin group, so the agent can create content and work with Remote Sync.3. We also recommend turning off the sample content and usage analytics, so they don't pollute the data model. If you're using a [docker compose file](../installation-and-operation/running-metabase-on-docker.md), add these [environment variables](../configuring-metabase/environment-variables.md):-```MB_LOAD_SAMPLE_CONTENT: "false"MB_INSTALL_ANALYTICS_DATABASE: "false"@@ -56,271 +57,89 @@ MB_INSTALL_ANALYTICS_DATABASE: "false"### Set up a repository to version your YAML files-1. Initialize a new repo.-2. Add a `.gitignore` file and add `.metabase/` and `.env`.-3. Add the following to your `.env`:--```- METABASE_URL={your-metabase-url}- METABASE_API_KEY={your-api-key}-```--### Add agent skills to your repository--You should add the following skills to your agent so it has context it needs. If you commit the skills into `.claude/skills/` in your repo, for example, Claude loads them automatically whenever you run it from that directory.--- [**`metabase-representation-format` agent skill**](https://github.com/metabase/agent-skills/blob/main/skills/metabase-representation-format/SKILL.md): teaches the agent the representation format and ships the schema checker.-- [**`metabase-database-metadata` agent skill**](https://github.com/metabase/agent-skills/blob/main/skills/metabase-database-metadata/SKILL.md): fetches database metadata from your Metabase into an on-disk YAML tree the agent can read while editing.-- [**`metabase-semantic-checker` agent skill**](https://github.com/metabase/agent-skills/blob/main/skills/metabase-semantic-checker/SKILL.md) (optional): runs Metabase's semantic checker in Docker to catch referential and query errors the schema check doesn't.--### Download the database metadata--Invoke the `metabase-database-metadata` skill and ask your agent to fetch the database metadata. The agent will:--- Check that `.env` exists. If it doesn't, the agent will prompt you to create it.-- Verify that `.env` and `.metabase/` are in `.gitignore`, asking before adding them.-- Fetch `/api/database/metadata` into `.metabase/metadata.json` (raw API response; can be several GB on large warehouses).-- Extract a diff-friendly YAML tree to `.metabase/databases/<database>/schemas/<schema>/tables/<table>.yaml` by running `npx @metabase/database-metadata extract-metadata`.--The agent can use the YAML extracted to your `.metabase` directory while creating and editing new questions and dashboards in YAML. That way your agent can refer to real column names, field types, and foreign-key relationships without making live API calls (which would be much slower).--To refresh this database metadata, just ask your agent to re-fetch it.--## Example workflows--The workflows below both assume you've completed the [Initial setup](#initial-setup).--### Example prompts--Once your repo has the agent skills and a baseline of content, prompt the agent with a structured request:+Create a new git repository for your Metabase content. You'll point Remote Sync at this repo when you configure it, and clone the repo locally to review changes and open pull requests.-```-Use the metabase-representation-format and metabase-database-metadata skills to create new YAML files in this directory:+### Install and authenticate the Metabase CLI-1. Create a new dashboard called "Support overview" in collections/main/.-2. Add questions showing total ticket volume, open tickets, and average satisfaction rating.-```--Or, depending on how capable your model is, try a more open-ended request:+Install the [Metabase CLI](../installation-and-operation/metabase-cli.md) globally:```-Use the metabase-representation-format and metabase-database-metadata skills to analyze our support data. Look at the tickets, customers, and interactions tables, and create a dashboard that gives an overview of our team's support workload.+npm install -g @metabase/cli```-The agent will read the representation format spec, check existing files for local conventions, consult `.metabase/databases/` for real column names, and write new YAML.--## Example workflow with Remote Sync--### 1. Configure Remote Sync on both instances--In your development Metabase, configure [Remote Sync in Read-write mode](../installation-and-operation/remote-sync.md#setting-up-remote-sync) pointed at your repo. In production, configure a second Metabase in Read-only mode pointed at the same repo.--### 2. Create a branch from the Metabase UI--Switch branches in Metabase, as the Metabase UI is the source of truth for which branch the development instance pushes to and pulls from.--In your development Metabase, click the **branch dropdown** at the top and [create a new branch](../installation-and-operation/remote-sync.md#creating-a-branch) for your work, like `feature/support-dashboard`.--### 3. Push existing content to seed the repo--Click the up arrow (**push**) icon to [commit and push](../installation-and-operation/remote-sync.md#committing-and-pushing-your-changes) your existing synced collections to the branch.+Then authenticate it against your development Metabase:-### 4. Clone the repo locally and check out the branch--```sh-git clone your-metabase-repo-cd your-metabase-repo-git checkout feature/support-dashboard```--### 5. Ask the agent to edit or create content--See [Example prompts](#example-prompts) above for prompt patterns to use here.--### 6. Validate the YAML files--Run the [schema check](#schema-check) after every batch of edits, and optionally run the [semantic check](#semantic-checker-for-deeper-validation) at the end of the session. See [Validating YAML files](#validating-yaml-files) below.--If anything fails, the agent should be able to fix the issue if you give it the error.--### 7. Commit and open a pull request--```sh-git add -A-git commit -m "Add support-overview dashboard"-git push origin feature/support-dashboard+mb auth login --url your-metabase-url-here```-Open a pull request so your team can review the YAML diff.--### 8. Pull the branch into your development Metabase--Click the **pull** (down arrow) icon in your development Metabase to load the agent's changes. Verify the dashboard renders correctly and the questions return expected results.+Authenticate with the API key you created in your Metabase instance.-### 9. Merge the PR so production picks up the changes+### Add the agent skill-If you've enabled [auto-sync](../installation-and-operation/remote-sync.md#pulling-changes-automatically), your production Metabase (in Read-only mode) will pull the new main branch automatically on its next interval. Otherwise, trigger a pull from production manually.+Add the [`/metabase-cli` skill](https://github.com/metabase/agent-skills/tree/main/skills/metabase-cli) to your agent so it knows how to use the CLI to create content directly in your Metabase.-## Example workflow with import and export endpoints+## Example prompts-### 1. Clone the empty repo and create a branch+These examples assume you've completed the [Initial setup](#initial-setup). Run the `/metabase-cli` skill and give the agent a structured request. The agent will run CLI commands to create the content directly in your development Metabase:-```sh-git clone https://github.com/your-org/your-repo.git-cd your-repo-git checkout -b feature/support-dashboard```--### 2. Export existing content to seed the repo--The agent does better work when the repo already holds your current Metabase content, so it can see real examples of the Representation Format and your collection conventions. Run the [serialization export](../installation-and-operation/serialization.md#serialization-workflow-example) from inside the clone:--```sh-curl \- -H 'X-API-Key: YOUR_API_KEY' \- -X POST 'https://your-metabase-url/api/ee/serialization/export?data_model=false' \- -o metabase_data.tgz-tar -xzf metabase_data.tgz-```--Set `data_model=false` to keep the export small. The agent should get its metadata from the `metabase-database-metadata` skill instead. For more on export options, see [Serialization](../installation-and-operation/serialization.md).--Commit the extracted YAML so you have a baseline to revert to if the agent goes off the rails.--### 3. Ask the agent to edit or create content--See [Example prompts](#example-prompts) above for prompt patterns to use here.--### 4. Validate the YAML files--Run the [schema check](#schema-check) after every batch of edits, and optionally run the [semantic check](#semantic-checker-for-deeper-validation) at the end of the session. See [Validating YAML files](#validating-yaml-files) below.--### 5. Commit and open a pull request--```sh-git add -A-git commit -m "Add support-overview dashboard"-git push origin feature/support-dashboard+/metabase-cli Create a new dashboard called "Support overview". Add questions showing total ticket volume, open tickets, and average satisfaction rating.```-Then open open a pull request so your team can review the YAML diff.--### 6. Import the YAML into your development Metabase--Re-bundle the YAML and import it:+Or, depending on how capable your model is, try a more open-ended request:-```sh-tar -czf metabase_data.tgz metabase_data-curl -X POST \- -H 'X-API-Key: YOUR_API_KEY' \- -F 'file=@metabase_data.tgz' \- 'https://your-metabase-url/api/ee/serialization/import' \- -o -```--The `-o -` flag writes the import response to stdout, so you can see whether the import succeeded and check any warnings. You can ask the agent to generate `export.sh` and `import.sh` wrappers so you're running a single command each time.--Verify the dashboard renders correctly and the questions return expected results.--### 7. Repeat the import against production--Once you're confident in the changes, run the same `tar` + `curl` import against your production Metabase, using its API key.--## Undoing the agent's changes--If you want to undo the agent's changes, use `git` to revert your YAML files to the last known-good commit before pushing or re-importing.--If you're using Remote Sync, don't try to fix things by re-pushing from Metabase: Metabase's push only reflects its current state and won't delete any new files the agent created locally.--## Validating YAML files--Run both checks locally before pushing. The same checks belong in CI — see [CI example](#ci-example) below.--### Schema check--You can run a quick schema check:--```sh-npx --yes @metabase/representations validate-schema+/metabase-cli Analyze our support data. Look at the tickets, customers, and interactions tables, and build a dashboard that gives an overview of our team's support workload.```-The check validates the shape of every YAML file against the Representation Format spec. The `metabase-representation-format` skill should run this check for you automatically after the agent makes any edits.+The agent writes the CLI commands and creates the questions and dashboard for you—just describe what you want.-### Semantic checker for deeper validation+## Example workflow-> The semantic checker is only available in the Pro/Enterprise plans.+### 1. Configure Remote Sync on both Metabase instances-The **semantic checker** catches things like references to tables that don't exist or columns the agent invented.+In your development Metabase, configure [Remote Sync in Read-write mode](../installation-and-operation/remote-sync.md#setting-up-remote-sync) pointed at your repo.-What it validates beyond schema:+Set up Remote Sync in your production Metabase in Read-only mode pointed at the same repo.-- Cross-entity references: `collection_id`, `dashboard_id`, `parent_id`, snippet names, transform tags, card embeddings.-- MBQL query compilation: `source-table`, field references, joins, segments, measures, expressions.-- Native-query references: tables, columns, and snippets named in SQL.--If you've installed the `metabase-semantic-checker` skill, just ask the agent to run the semantic checker; the skill picks the right image, passes the right flags, and summarizes the findings.+### 2. Create a branch from the Metabase UI-You can manually run the semantic checker via Docker like so:+Switch branches in Metabase, as the Metabase UI is the source of truth for which branch the development instance pushes to and pulls from.-```sh-docker pull metabase/metabase-enterprise:latest+In your development Metabase, click the **branch dropdown** at the top and [create a new branch](../installation-and-operation/remote-sync.md#creating-a-branch) for your work, like `feature/support-dashboard`.-docker run --rm \- -v "$PWD:/workspace" \- --entrypoint "" \- -w /app \- metabase/metabase-enterprise:latest \- java -jar metabase.jar \- --mode checker \- --export /workspace \- --schema-dir /workspace/.metabase/metadata.json \- --schema-format concise-```+### 3. Ask the agent to create content-Match the image tag (`:latest`) to your Metabase build.+Run the `/metabase-cli` skill and prompt the agent to build your questions and dashboards. The agent creates the content directly in your development Metabase.-### CI example+### 4. Verify the content in your development Metabase-You can hook the schema check into GitHub Actions so your team catches problems on the PR, before anyone pulls the changes into Metabase:+Open your development Metabase and confirm the dashboard renders correctly and the questions return expected results. Make any changes you want, either in the UI or via the agent.-```yaml-# .github/workflows/schema-check.yml-name: Schema Check+### 5. If you make any changes in your Metabase, push the new content to the branch-on:- push:- branches: [main]- pull_request:+To commit the work, [push the change from your Metabase](../installation-and-operation/remote-sync.md#committing-and-pushing-your-changes). If you don't push your changes from your Metabase to your repo, you could lose work the next time you pull changes into that Metabase.-jobs:- schema-check:- runs-on: ubuntu-latest- steps:- - uses: actions/checkout@v6+### 6. Open a pull request- - uses: actions/[email protected]- with:- node-version: "20"+Open a pull request so your team can review the YAML diff. They can also use Remote Sync to pull the branch into a development Metabase and see the changes live.- - name: Validate representation YAML files- run: npx --yes @metabase/representations validate-schema-```+### 7. Merge the PR so production picks up the changes-For the semantic check, add a second workflow that fetches `.metabase/metadata.json` from your Metabase and then runs the Docker command above against the checkout. If you run the semantic check in more than one workflow (for example, a semantic check and per-PR preview environments), you should probably factor the database metadata fetch to run and cache once a day so you don't hit the API on every push.+If you've enabled [auto-sync](../installation-and-operation/remote-sync.md#pulling-changes-automatically), your production Metabase (in Read-only mode) will pull the new main branch automatically on its next interval.-## Deleting content+## Undoing the agent's changes-Since imports and exports _don't_ delete content, you'll need to delete content in the Metabase application itself, then update the YAML files as well.+Since the agent uses the CLI to create content directly in Metabase, to undo changes, you could either:-1. Delete the content in your production Metabase (in the app's UI).-2. Push (with Remote Sync) or re-export (without) so the change is reflected in the repo.-3. Commit the deletion. That way Metabase won't recreate the deleted items the next time it pulls.+- **Revert to a commit and pull changes**: Revert to a previous commit, and pull that commit into your Metabase. Pulling that commit would overwrite any changes you or the agent had made.+- **Update Metabase and push changes**: Edit or archive the content in your Metabase (either manually or via an agent with the CLI), and push the changes to your repo. You may want to pick this approach if you want to keep some, but not all of the changes made by the AI, or if you've made additional changes via Metabase's handy UI that you want to keep.## Further reading- [Remote Sync](../installation-and-operation/remote-sync.md)-- [Serialization](../installation-and-operation/serialization.md)+- [Metabase CLI](../installation-and-operation/metabase-cli.md)- [Metabase Representation Format](https://github.com/metabase/representations)-- [Metabase Database Metadata Format](https://github.com/metabase/database-metadata)- [Agent skills](https://github.com/metabase/agent-skills)-- [MCP server](./mcp.md) — for agents that need live metadata lookups outside the file-based workflow.+- [MCP server](./mcp.md): for agents that need live metadata lookups outside the file-based workflow.diff --git a/docs/ai/overview.md b/docs/ai/overview.mdindex 07f21dfd1cd..45d512c0e83 100644--- a/docs/ai/overview.md+++ b/docs/ai/overview.md@@ -50,7 +50,7 @@ See [full docs for MCP server](mcp.md).**Best for: developers creating stuff that other people will use.**-Use a coding agent like Claude Code to understand your database's metadata, generate Metabase content as YAML files locally, verify the schema, then sync and import the generated content into your production Metabase. Sky's the limit on what you can accomplish.+Use a coding agent like Claude Code with the Metabase CLI to create Metabase content directly in a development instance, then version that content as YAML files and pull the changes into your production Metabase.**Controls**: Only admins can sync content to Metabase instances.diff --git a/docs/installation-and-operation/commands.md b/docs/installation-and-operation/commands.mdindex 13618391431..0d10d6933c0 100644--- a/docs/installation-and-operation/commands.md+++ b/docs/installation-and-operation/commands.md@@ -1,11 +1,13 @@----title: Metabase CLI-description: CLI commands for managing your Metabase instance, including database migrations, serialization, and administrative tasks.+title: Metabase JAR commands+description: Commands built into the Metabase JAR for managing your instance, including database migrations, serialization, and administrative tasks.----# Metabase CLI+# Metabase JAR commands-Metabase ships with some handy CLI commands for administration, maintenance, and automation tasks. These commands let you manage your Metabase instance, migrate databases, handle serialization, and generate documentation.+> Looking for the `mb` command-line client that drives a Metabase instance over its API, on its own or through an AI agent? Check out the [Metabase CLI](./metabase-cli.md).++Metabase ships with some handy commands for administration, maintenance, and automation tasks. These commands run on the server, built into the Metabase JAR, and let you manage your Metabase instance, migrate databases, handle serialization, and generate documentation.To view a list of commands, run the Metabase jar followed by `help`.@@ -15,7 +17,6 @@ java --add-opens java.base/java.nio=ALL-UNNAMED -jar metabase.jar helpMetabase will print out the help text for available commands.-## `api-documentation`Generate an HTML file and a JSON file for Scalar docs for the Metabase API.@@ -118,4 +119,3 @@ Open an SQL shell for the Metabase H2 DB:```shjava -cp metabase.jar org.h2.tools.Shell -url jdbc:h2:/path/to/metabase.db```-diff --git a/docs/installation-and-operation/metabase-cli.md b/docs/installation-and-operation/metabase-cli.mdnew file mode 100644index 00000000000..7b3cb9ca136--- /dev/null+++ b/docs/installation-and-operation/metabase-cli.md@@ -0,0 +1,77 @@+---+title: Metabase CLI+summary: "The Metabase CLI (mb) is a command-line client that drives a Metabase instance over its API. Use it yourself, or hand it to an AI agent with the metabase-cli skill."+---++# Metabase CLI++The Metabase CLI (`mb`) is a command-line client for Metabase. `mb` authenticates against a Metabase instance with an API key and lets you or an AI agent read and write content like questions, dashboards, collections, and transforms over the Metabase API.++> Looking for the commands built into the Metabase JAR, like `migrate` or `load-from-h2`? Check out [Metabase JAR commands](./commands.md).++## Requirements++- Node.js, to install the CLI from npm.+- A Metabase instance on version 58 or later.+- An [API key](../people-and-groups/api-keys.md#create-an-api-key) to authenticate the CLI against your instance.+- A Pro or Enterprise plan for some command groups. For example, `git-sync` needs the premium [Remote sync](./remote-sync.md) feature.++## Install the CLI++```+npm install -g @metabase/cli+```++The binary is `mb`.++For commands, run:++```+mb --help+```++## Authenticate the CLI++Log in once per Metabase instance.++```+mb auth login --url https://metabase.example.com+```++The CLI prompts for an API key, or reads it from the `METABASE_API_KEY` environment variable or stdin. To create a key, see [API keys](../people-and-groups/api-keys.md#create-an-api-key).++Credentials are stored per profile, so you can manage more than one Metabase (like dev and prod Metabases):++```+mb auth login --profile prod --url https://prod.example.com+mb auth list+```++Add `--profile <name>` to any command to run it against that instance.++## Use the CLI with an AI agent++The CLI is built to be driven by an AI coding agent like Claude Code. Instead of running commands yourself, you install a skill and describe what you want in plain language; the agent works out the commands.++### The metabase-cli skill++The [metabase-cli skill](https://github.com/metabase/agent-skills/tree/main/skills/metabase-cli) teaches your agent the CLI's conventions. Once installed, you can run:++```+/metabase-cli Create a dashboard summarizing this month's signups by plan.+```++And your agent will go to work, creating content directly in your Metabase via the `mb` CLI.++## Use the CLI for agent-driven development++Pair the CLI with version control to build content with an agent in a development Metabase, commit the changes, and pull the changes into your production Metabase. Check out [Agent-driven development](../ai/file-based-development.md).++## Further reading++- [@metabase/cli on npm](https://www.npmjs.com/package/@metabase/cli)+- [Agent-driven development](../ai/file-based-development.md)+- [Agent skills](https://github.com/metabase/agent-skills)+- [Remote sync](./remote-sync.md)+- [Serialization](./serialization.md)+- [Metabase JAR commands](./commands.md)diff --git a/docs/installation-and-operation/start.md b/docs/installation-and-operation/start.mdindex 97a9fd929b4..c8f975ad334 100644--- a/docs/installation-and-operation/start.md+++ b/docs/installation-and-operation/start.md@@ -34,6 +34,10 @@ Try out your changes before deploying to production.Export your questions, dashboards, and more as YAML files that you can check into version control.+## [Metabase CLI](./metabase-cli.md)++A command-line client that drives a Metabase instance over its API, on its own or through an AI agent.+## [Monitoring your Metabase](./monitoring-metabase.md)Monitor your Metabase with JMX.@@ -54,9 +58,9 @@ We collect anonymous Metabase usage data to improve the product. We don't collecNotes on Metabase's accessibility.-## [Command line commands](./commands.md)+## [Metabase JAR commands](./commands.md)-Metabase's CLI.+Administrative commands built into the Metabase JAR.---diff --git a/resources/commands/header.md b/resources/commands/header.mdindex 1309851520f..425936038ea 100644--- a/resources/commands/header.md+++ b/resources/commands/header.md@@ -1,11 +1,13 @@----title: Metabase CLI-description: CLI commands for managing your Metabase instance, including database migrations, serialization, and administrative tasks.+title: Metabase JAR commands+description: Commands built into the Metabase JAR for managing your instance, including database migrations, serialization, and administrative tasks.----# Metabase CLI+# Metabase JAR commands-Metabase ships with some handy CLI commands for administration, maintenance, and automation tasks. These commands let you manage your Metabase instance, migrate databases, handle serialization, and generate documentation.+> Looking for the `mb` command-line client that drives a Metabase instance over its API, on its own or through an AI agent? Check out the [Metabase CLI](./metabase-cli.md).++Metabase ships with some handy commands for administration, maintenance, and automation tasks. These commands run on the server, built into the Metabase JAR, and let you manage your Metabase instance, migrate databases, handle serialization, and generate documentation.To view a list of commands, run the Metabase jar followed by `help`.