PRDCT-487: file the Getting Started side trips in their own sections

CLAUDE.md

@@ -60,7 +60,7 @@ The full repo guide — stack, structure, commands, and conventions — lives in
 - Conserve content: every removed paragraph → a new home or a logged deletion
   (CONSERVATION-REPORT + MISSING list). Never drop silently.
 - Current guardrails: don't assert E2B as hosting; don't say Streamlit is retired;
-  don't document pricing / in-app Kai-toggle / semantic layer.
+  don't document pricing / in-app Kai-toggle.
 - Stakeholder likes/dislikes + the paste-ready prompt block live in the local
   `apps-section/` bundle (gitignored), not here.
 

_data/navigation.yml

@@ -50,41 +50,6 @@ items:
       - url: /getting-started/next-steps/
         title: Where to Go Next
 
-      - title: Going Further
-        items:
-          - url: /getting-started/load/googlesheets/
-            title: Load from Google Sheets
-
-          - url: /getting-started/load/database/
-            title: Load from a Database
-
-          - url: /getting-started/transform/workspace/
-            title: Use a Workspace
-
-          - url: /getting-started/ad-hoc/
-            title: Ad-Hoc Data Analysis
-
-          - url: /getting-started/branches/
-            title: Development Branches
-            items:
-              - url: /getting-started/branches/prepare-tables/
-                title: Prepare Tables
-
-              - url: /getting-started/branches/prepare-files/
-                title: Prepare Files
-
-              - title: Tables in Branch
-                url: /getting-started/branches/tables-in-branch/
-
-              - title: Files in Branch
-                url: /getting-started/branches/files-in-branch/
-
-              - url: /getting-started/branches/project-diff/
-                title: Project Diff
-
-              - title: Merge to Production
-                url: /getting-started/branches/merge-to-production/
-
   - url: /kai/
     title: Kai - AI Assistant
     items:
@@ -343,7 +308,7 @@ items:
                 title: FTP
 
               - url: /components/extractors/storage/google-drive/
-                title: Google Drive
+                title: Google Sheets
 
               - url: /components/extractors/storage/http/
                 title: HTTP
@@ -636,6 +601,27 @@ items:
           - url: /components/branches/merge-requests/
             title: Merge Requests
 
+          - url: /components/branches/tutorial/
+            title: Branches Tutorial
+            items:
+              - url: /components/branches/tutorial/prepare-tables/
+                title: Prepare Tables
+
+              - url: /components/branches/tutorial/prepare-files/
+                title: Prepare Files
+
+              - url: /components/branches/tutorial/tables-in-branch/
+                title: Tables in Branch
+
+              - url: /components/branches/tutorial/files-in-branch/
+                title: Files in Branch
+
+              - url: /components/branches/tutorial/project-diff/
+                title: Project Diff
+
+              - url: /components/branches/tutorial/merge-to-production/
+                title: Merge to Production
+
       - url: /components/ip-addresses/
         title: IP Addresses
 
@@ -798,6 +784,12 @@ items:
       - url: /workspace/snowflake-workspaces-access-changes/
         title: Snowflake Workspaces Access Changes
 
+      - url: /workspace/create/
+        title: Create a Workspace
+
+      - url: /workspace/ad-hoc-analysis/
+        title: Ad-Hoc Data Analysis
+
       - url: /workspace/sql-editor/
         title: SQL Editor
 
@@ -877,6 +869,9 @@ items:
       - url: /ai/mcp-server/
         title: MCP Server
 
+      - url: /ai/semantic-layer/
+        title: Semantic Layer
+
   - url: /extend/
     title: Extending Keboola
     items:

src/content/docs/ai/ai-kit/index.md

@@ -7,9 +7,9 @@ slug: 'ai/ai-kit'
 
 AI Kit is a plugin marketplace for AI coding assistants that provides specialized agents, commands, and workflows for Keboola development. It helps developers build Keboola components, data apps, and maintain code quality using AI-powered tools.
 
-AI Kit is designed for developers who use AI coding assistants like Claude Code to work with Keboola projects. It provides three specialized plugins that cover different aspects of Keboola development, from general code quality to building production-ready components and data applications.
+AI Kit is designed for developers who use AI coding assistants like Claude Code to work with Keboola projects. It provides seven plugins that cover different aspects of Keboola development, from building production-ready components and data applications to driving your projects from the terminal and modelling your semantic layer.
 
-The toolkit includes specialized AI agents that understand Keboola's architecture, best practices, and development patterns. These agents can help you create new components from scratch, implement configuration schemas, build Streamlit data apps, review code for security issues, and automate common development workflows.
+The toolkit includes specialized AI agents that understand Keboola's architecture, best practices, and development patterns. These agents can help you create new components from scratch, implement configuration schemas, build data apps, review a project for SQL and security problems, and automate common development workflows.
 
 ## Installation
 
@@ -22,36 +22,26 @@ To install AI Kit, run the following command in your AI coding assistant:
 After installation, enable the plugins you need:
 
 ```bash
-/plugin install developer
-/plugin install component-developer
-/plugin install dataapp-developer
+/plugin install component-developer@keboola-claude-kit
+/plugin install dataapp-developer@keboola-claude-kit
+/plugin install kbagent@keboola-claude-kit
+/plugin install keboola-cli@keboola-claude-kit
+/plugin install keboola-git@keboola-claude-kit
+/plugin install powerbi-to-sl@keboola-claude-kit
+/plugin install sl-toolkit@keboola-claude-kit
 ```
 
-## Available Plugins
-
-### Developer Plugin
-
-The Developer Plugin provides a comprehensive toolkit for code quality, security analysis, and workflow automation. It includes four specialized agents and a PR creation command.
-
-**Agents:**
-
-The **Code Reviewer** (`@code-reviewer`) is an expert code reviewer that checks for bugs, logic errors, security vulnerabilities, and project guidelines compliance. It uses confidence-based filtering to report only high-priority issues.
+`keboola-claude-kit` is the marketplace name this repository publishes, and the one to install Keboola plugins from.
 
-The **Security Agent** (`@code-security`) performs cross-language security analysis for Python, Go, PHP, JavaScript, and other languages. It integrates with automated security scanners and identifies CWE/CVE vulnerabilities with actionable fixes.
+<!-- Plugin list, names and versions read from keboola/ai-kit .claude-plugin/marketplace.json on 2026-08-26: component-developer 3.3.2, dataapp-developer 1.5.1, kbagent 0.91.0, keboola-cli 1.1.1, keboola-git 1.0.1, powerbi-to-sl 1.0.0, sl-toolkit 3.0.0. There is no `developer` plugin. -->
 
-The **Code Mess Detector** (`@code-mess-detector`) analyzes code written during rapid prototyping for common quality issues like inconsistent naming, missing error handling, code duplication, and dead code. It generates detailed reports for systematic cleanup.
-
-The **Code Mess Fixer** (`@code-mess-fixer`) systematically applies fixes based on Code Mess Detector reports, working through issues by priority and tracking progress.
-
-**Commands:**
-
-The `/create-pr` command analyzes your changes and creates a pull request with an AI-generated title and description, following conventional commit format.
+## Available Plugins
 
-**Integrations:**
+### kbagent Plugin
 
-The plugin includes Linear MCP integration for issue tracking and project management, and auto-installs team-wide permission settings for safe git operations.
+The kbagent Plugin teaches an AI client the [kbagent CLI](/cli/), so the assistant can run jobs, read configurations and search across your Keboola projects. It adds a `/keboola` command backed by a `keboola-expert` subagent, a skill that understands plain-language asks like "set up kbagent for my Keboola project", and in Claude Code a `/kbagent:setup` command that installs and connects the CLI in one step.
 
-[View Developer Plugin Documentation on GitHub](https://github.com/keboola/ai-kit/tree/main/plugins/developer)
+The plugin's source lives in the CLI repo at [`plugins/kbagent`](https://github.com/keboola/cli/tree/main/plugins/kbagent); this marketplace publishes it from there through a `git-subdir` source. Claude Code, Claude Desktop, Cursor, VS Code and the ChatGPT app can all install it, each by its own route. See [kbagent with AI agents](/cli/for-agents/) for the steps per client.
 
 ### Component Developer Plugin
 
@@ -81,6 +71,30 @@ The component should support incremental loads based on a timestamp field.
 
 [View Component Developer Plugin Documentation on GitHub](https://github.com/keboola/ai-kit/tree/main/plugins/component-developer)
 
+### Keboola CLI Plugin
+
+The Keboola CLI Plugin is a project management and review toolkit built on the older `kbc` sync CLI. It ships a ten-agent review team that analyses a project for SQL quality, security, performance, financial logic, and template readiness. It is a separate tool from the kbagent plugin above. If you want the agent interface to your projects, install `kbagent`.
+
+[View Keboola CLI Plugin Documentation on GitHub](https://github.com/keboola/ai-kit/tree/main/plugins/keboola-cli)
+
+### Keboola Git Plugin
+
+The Keboola Git Plugin works with Keboola-managed Git (Forgejo) repositories for Python/JS data apps, which can host their source in Keboola rather than GitHub. It provisions repositories, mints push credentials, and copies source between GitHub and Keboola git through the kbagent CLI. It also carries the 15 MB push cap and the build-at-deploy workaround it forces.
+
+[View Keboola Git Plugin Documentation on GitHub](https://github.com/keboola/ai-kit/tree/main/plugins/keboola-git)
+
+### Semantic Layer Toolkit
+
+The `sl-toolkit` plugin inspects, validates, and builds semantic layer models through the metastore API. `/sl-show` lists a model's datasets, metrics, and relationships, `/sl-validate` checks it for phantom fields and dangling references, and `/sl-build` walks a greenfield model from schema discovery to push. Adding and editing model objects works conversationally, with no slash command.
+
+[View Semantic Layer Toolkit Documentation on GitHub](https://github.com/keboola/ai-kit/tree/main/plugins/sl-toolkit)
+
+### Power BI to Semantic Layer
+
+The `powerbi-to-sl` plugin migrates an existing Microsoft Power BI semantic model into a Keboola semantic layer model, translating tables, columns, measures, and relationships. It is the brownfield companion to `sl-toolkit`, which generates a new model instead.
+
+[View Power BI to Semantic Layer Documentation on GitHub](https://github.com/keboola/ai-kit/tree/main/plugins/powerbi-to-sl)
+
 ### Data App Developer Plugin
 
 The Data App Developer Plugin is a specialized toolkit for building production-ready Streamlit data apps for Keboola deployment. It features a systematic validate, build, and verify workflow that ensures features work correctly the first time.
@@ -108,7 +122,7 @@ The agent will automatically validate the schema, query distinct values, create
 
 ## Best Practices
 
-When using AI Kit, start with the appropriate plugin for your task. Use the Developer Plugin for general code quality and security reviews, the Component Developer Plugin when building new Keboola components or adding features to existing ones, and the Data App Developer Plugin when creating or modifying Streamlit data apps.
+When using AI Kit, start with the appropriate plugin for your task. Use the Component Developer Plugin when building new Keboola components or adding features to existing ones, the Data App Developer Plugin when creating or modifying data apps, the kbagent Plugin to drive your projects from the terminal, and the Semantic Layer Toolkit when modelling metrics.
 
 For component development, always follow the two-PR workflow strategy: create a base PR with the cookiecutter-generated structure, then a separate implementation PR with your custom logic. This prevents premature CI/CD triggers.
 

src/content/docs/ai/index.md

@@ -27,6 +27,11 @@ A plugin marketplace for AI coding assistants that provides specialized agents,
 Model Context Protocol server integration for seamless communication between AI agents and your data infrastructure on Keboola.
 [Learn more about MCP Server →](/ai/mcp-server/)
 
+### Semantic Layer
+
+Describe your data in business terms — datasets, metrics, relationships, glossary terms, and business rules — so AI assistants understand what your data means, not just how it is stored.
+[Learn more about the Semantic Layer →](/ai/semantic-layer/)
+
 ### Machine-Readable API Index
 
 Every Keboola stack publishes a machine-readable index of its APIs at `https://api.<stack>/apis.json` (for example, [`https://api.keboola.com/apis.json`](https://api.keboola.com/apis.json)). The index lists each service with its base `apiUrl` and a link to its OpenAPI specification (`openApiSpecUrl`) — a convenient overview for agentic usage: AI agents, MCP servers, and other tooling that discovers Keboola's APIs programmatically.

src/content/docs/ai/mcp-server/index.md

@@ -249,7 +249,7 @@ Don't worry about remembering command names — your AI client handles that. Jus
 - **Components & Transformations** – Create, edit, and launch them with natural language.  
 - **Storage** – Browse, edit, and document buckets, tables, and columns.  
 - **SQL** – Run and manage SQL queries.  
-- **Semantic layer** – Explore the project's semantic models and validate queries against them.  
+- **Semantic layer** – Explore the project's semantic models and validate queries against them. See [Semantic Layer](/ai/semantic-layer/).  
 - **Jobs** – Start, monitor, and debug execution flows.  
 - **Flows** – Create and manage flows (including conditional flows) that orchestrate your components.  
 - **Data Apps** – Create, deploy, and manage Streamlit and Python/JS data apps.  
@@ -301,13 +301,13 @@ The following tools are classified as read-only (they do not modify data). The l
 |----------|-------|
 | Components | `get_configs`, `get_components`, `get_config_examples`, `run_sync_action` |
 | Flows | `get_flows`, `get_flow_examples`, `get_flow_schema` |
-| Storage | `get_buckets`, `get_tables` |
+| Storage | `get_buckets`, `get_shared_buckets`, `get_tables` |
 | SQL | `query_data` |
 | Semantic | `get_semantic_context`, `get_semantic_schema`, `search_semantic_context`, `validate_semantic_query` |
 | Data Apps | `get_data_apps` |
 | Jobs | `get_jobs` |
 | Search | `search`, `find_component_id` |
-| Project | `get_project_info` |
+| Project | `get_accessible_projects`, `get_project_info`, `set_project_scope` |
 | Documentation | `docs_query` |
 
 ### Examples
@@ -436,7 +436,7 @@ The primary way to run the server locally without Docker is by using `uv` or `uv
    * `KBC_STORAGE_API_URL`: Your Keboola instance API URL (e.g., `https://connection.keboola.com` or `https://connection.YOUR_REGION.keboola.com`).
    * `KBC_BRANCH_ID` (optional): a development branch ID to scope operations to; defaults to the production branch.
 
-   Refer to the [Keboola Tokens](/management/project/tokens/) and [Keboola workspace manipulation](/getting-started/transform/workspace/) for detailed instructions on obtaining these values.
+   Refer to the [Keboola Tokens](/management/project/tokens/) and [Keboola workspace manipulation](/workspace/create/) for detailed instructions on obtaining these values.
 
    **1.1. Additional Setup for BigQuery Users**  
    If your Keboola project uses BigQuery as its backend, you will also need to set up the `GOOGLE_APPLICATION_CREDENTIALS` environment variable. This variable should point to the JSON file containing your Google Cloud service account key that has the necessary permissions to access your BigQuery data.
@@ -579,7 +579,7 @@ For detailed instructions and SDKs for building your own MCP client, refer to th
 
 ## Advanced Setup Options
 These methods are for developers or specific use cases (e.g., testing, contributing to the MCP server).
-Prefer a terminal or want to give an agent sandboxed, multi-project control? See the [kbagent CLI](/cli/) — it can also call MCP tools via `kbagent tool`. For dev environments or contributing to the MCP Server, check out the [MCP GitHub repo](https://github.com/keboola/mcp-server).
+Prefer a terminal or want to give an agent sandboxed, multi-project control? See the [kbagent CLI](/cli/). For dev environments or contributing to the MCP Server, check out the [MCP GitHub repo](https://github.com/keboola/mcp-server).
 
 
 ## Support and Feedback