@kvasar/openclaw-storyblok-plugin 0.1.1 → 0.1.3

package/openclaw.plugin.json

@@ -1,9 +1,9 @@
 {
   "id": "openclaw-storyblok",
-  "kind": "tool",
   "name": "Storyblok Integration",
-  "version": "0.1.1",
+  "version": "0.1.3",
   "description": "Provides tools to interact with Storyblok CMS via Management API and Delivery API. Supports stories, components, and space management.",
+  "skills": ["skills"],
   "configSchema": {
     "type": "object",
     "additionalProperties": false,

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@kvasar/openclaw-storyblok-plugin",
-  "version": "0.1.1",
+  "version": "0.1.3",
   "description": "OpenClaw plugin — interact with Storyblok CMS via Management API and Delivery API",
   "type": "module",
   "main": "./index.ts",

package/skills/README.md

@@ -0,0 +1,54 @@
+# Storyblok Skill Package
+
+This package provides a comprehensive set of skills for working with Storyblok CMS from OpenClaw agents.
+
+## Quick Start
+
+1. Ensure the `openclaw-storyblok` plugin is installed and configured.
+2. In a session, activate this skill: `skill storyblok`
+3. Use the appropriate prompt for your workflow:
+   - Content Query: `use prompt content-query`
+   - Content Management: `use prompt content-management`
+   - Page Generation: `use prompt page-generation`
+   - Sync Workflow: `use prompt sync-workflow`
+
+Or invoke the skill directly and follow the guidance in `SKILL.md`.
+
+## Contents
+
+- `SKILL.md` – Full skill definition, tool reference, and workflow guides.
+- `prompts/` – Prompt templates for each workflow.
+- `workflows/` – Step-by-step workflow descriptions.
+- `examples/` – Concrete usage examples.
+
+## Tools Provided by the Plugin
+
+- `storyblok_get_space`
+- `storyblok_get_story`
+- `storyblok_list_stories`
+- `storyblok_get_components`
+- `storyblok_create_story`
+- `storyblok_update_story`
+- `storyblok_publish_story`
+- `storyblok_unpublish_story`
+
+## Additional Tools (used by skill)
+
+- `sessions_spawn`
+- `read`, `write`, `edit`
+- `git_status`, `git_add`, `git_commit`, `git_push`
+
+## Example Session
+
+```
+> skill storyblok
+Using skill: storyblok
+
+> what stories are in folder 123?
+Use storyblok_list_stories with folder_id=123
+storyblok_list_stories folder_id=123
+```
+
+## License
+
+Part of the OpenClaw plugin ecosystem.

package/skills/SKILL.md

@@ -0,0 +1,218 @@
+---
+name: storyblok
+description: Comprehensive skill for interacting with Storyblok CMS: query content, manage stories, generate pages, and sync with frontend repositories.
+allowed-tools:
+  - storyblok_get_space
+  - storyblok_get_story
+  - storyblok_list_stories
+  - storyblok_get_components
+  - storyblok_create_story
+  - storyblok_update_story
+  - storyblok_publish_story
+  - storyblok_unpublish_story
+  - sessions_spawn
+  - read
+  - write
+  - edit
+  - git_status
+  - git_add
+  - git_commit
+  - git_push
+---
+
+# Storyblok Integration Skill
+
+This skill package provides agents with the ability to work with **Storyblok CMS** using the `openclaw-storyblok` plugin. It covers four main workflows:
+
+1. **Content Query** – Retrieve spaces, stories, components, and listings.
+2. **Content Management** – Create, update, publish, and unpublish stories.
+3. **Page Generation** – Use AI-assisted reasoning to create new landing pages.
+4. **Sync Workflow** – Synchronize content or schemas between Storyblok and a frontend repository.
+
+## Prerequisites
+
+The Storyblok plugin must be configured with valid credentials:
+
+- `baseUrl` (default: https://api.storyblok.com)
+- `spaceId`
+- `managementToken` (required for write operations)
+- `previewToken` (optional for read-only operations)
+
+These are set when installing the plugin. Inside a session, the plugin config will be used automatically.
+
+## Core Concepts
+
+### Tools
+
+All tools are provided by the `openclaw-storyblok` plugin:
+
+| Tool | Purpose |
+|------|---------|
+| `storyblok_get_space` | Get space details |
+| `storyblok_get_story` | Fetch a story by ID/UUID/slug |
+| `storyblok_list_stories` | List stories with optional filters |
+| `storyblok_get_components` | Retrieve component schemas (blok types) |
+| `storyblok_create_story` | Create a new story or folder |
+| `storyblok_update_story` | Update an existing story |
+| `storyblok_publish_story` | Publish a draft story |
+| `storyblok_unpublish_story` | Unpublish a published story |
+
+### Parameter Notes
+
+- Most tools accept optional parameters; only required fields are mandatory.
+- For `storyblok_get_story`, you can specify `version` (draft/published) and `language`.
+- For `storyblok_list_stories`, you can filter by `folder_id`, `parent_id`, `status`, `tag`, `per_page`, `page`, `sort_by`, `direction`.
+- For management tools, `story_id` can be numeric ID or UUID.
+
+### Error Handling
+
+All tools return a standardized response:
+- Success: `{ content: [{type: "text", text: JSON}], metadata: {...} }`
+- Error: `{ content: [{type: "text", text: "❌ Storyblok: ..."}], isError: true }`
+
+Agents should check `isError` and display the error message to the user.
+
+## Workflow Guides
+
+### 1. Content Query
+
+Use the query tools to explore and retrieve data.
+
+**When to use what:**
+
+- Need space info? → `storyblok_get_space`
+- Need a specific story? → `storyblok_get_story` (provide `identifier`)
+- Need a list of stories with filters? → `storyblok_list_stories` (set filters)
+- Need to know available content block types? → `storyblok_get_components`
+
+**Example: List all published stories in the last week**
+
+```sh
+storyblok_list_stories status="published" per_page=100
+```
+
+**Example: Get a story by slug**
+
+```sh
+storyblok_get_story identifier="my-landing-page" version="published"
+```
+
+### 2. Content Management
+
+Use the management tools to create, modify, and publish content.
+
+**Workflow:**
+
+- To create a new story: `storyblok_create_story` with `title` and optional `content`, `folder_id`, etc.
+  - The content is a JSON object matching your component structure.
+  - You can use `storyblok_get_components` first to see available block types.
+- To edit a story: `storyblok_update_story` with `story_id` and fields to update.
+  - Specify `version="draft"` (default) or `"published"` to update a published version (use with caution).
+- To publish: `storyblok_publish_story` with `story_id`.
+- To unpublish: `storyblok_unpublish_story` with `story_id`.
+
+**Example: Create a draft story**
+
+```sh
+storyblok_create_story title="New Homepage" content={"title":"Welcome","body":"<p>Hello</p>"}
+```
+
+**Example: Publish a story**
+
+```sh
+storyblok_publish_story story_id="12345" publish_notes="Initial release"
+```
+
+### 3. Page Generation
+
+This workflow combines AI reasoning with Storyblok tools to produce complete pages.
+
+**Steps:**
+
+1. Clarify the page goal with the user (topic, tone, key messages).
+2. Fetch component schemas via `storyblok_get_components` to understand available block types.
+3. Use reasoning to design a JSON content structure that uses those components.
+4. Create the story with `storyblok_create_story`.
+5. Optionally publish immediately with `storyblok_publish_story`.
+
+The agent may also spawn a sub-session (using `sessions_spawn`) to generate the content with an AI model if needed.
+
+### 4. Sync Workflow
+
+Synchronize Storyblok content with a frontend repository. This workflow uses file operations and git.
+
+**Schema Sync:**
+
+- Fetch component schemas from Storyblok (`storyblok_get_components`).
+- Compare with local component definitions (e.g., in a frontend project's component library).
+- If schemas are out of date, update local files to match (or notify user).
+- Commit and push changes (`git_add`, `git_commit`, `git_push`).
+
+**Content Sync:**
+
+- Iterate through stories (`storyblok_list_stories`, `storyblok_get_story`).
+- Export each story to a file (e.g., JSON or Markdown) in the repository.
+- Commit changes.
+
+**Considerations:**
+
+- Always operate within the designated repository directory (use `workdir` context).
+- Use `git_status` to see changes before committing.
+- Write clear commit messages.
+- Do not push without user approval (for safety, the skill could be configured to require explicit confirmation).
+
+## Session Awareness
+
+This skill is designed to work in both main and sub-agent sessions. It can be used directly or spawned via `sessions_spawn` for parallel or isolated execution. When spawning, pass the same tool permissions to ensure access to Storyblok tools.
+
+## README Usage Examples
+
+### Quick Query
+
+```
+User: What stories are in folder 123?
+Agent: storyblok_list_stories folder_id=123
+```
+
+### Create and Publish
+
+```
+User: Create a blog post titled "Latest Updates" with body "We've improved performance."
+Agent: storyblok_create_story title="Latest Updates" content={"post_content":"We've improved performance."}
+Agent: storyblok_publish_story story_id="<new_id>"
+```
+
+### Generate a Landing Page
+
+```
+User: Generate a landing page for our summer sale.
+Agent: (fetches components, designs structure, creates story)
+storyblok_get_components
+storyblok_create_story title="Summer Sale 2026" content={...}
+storyblok_publish_story story_id="..."
+```
+
+### Sync Schemas
+
+```
+User: Sync Storyblok component schemas with our UI library.
+Agent: git_status
+Agent: storyblok_get_components > components.json
+Agent: (compare, update local files)
+Agent: git_add src/components/storyblok/
+Agent: git_commit -m "chore: update Storyblok component schemas"
+Agent: git_push
+```
+
+## Notes
+
+- Always handle errors gracefully and report to the user.
+- Avoid publishing without explicit user consent.
+- For large exports, use pagination (`per_page`, `page`) to stay within response limits.
+- Respect rate limits: Storyblok API may have limits; consider adding delays if many requests.
+
+## Troubleshooting
+
+- **Authentication errors**: Check plugin configuration (tokens, space ID).
+- **Tool not found**: Ensure the `openclaw-storyblok` plugin is installed and loaded in the OpenClaw environment.
+- **Permission errors**: The management token must have write permissions; preview token only for reads.

package/skills/examples/management.md

@@ -0,0 +1,33 @@
+# Content Management Examples
+
+## Example 1: Create a simple story
+
+```
+User: Create a new story titled "Q2 Results" with a single paragraph.
+Agent:
+storyblok_create_story title="Q2 Results" content={"body":"<p>Our quarterly results are strong...</p>"}
+```
+
+## Example 2: Update a story's title
+
+```
+User: Change the title of story 555 to "2026 Annual Report"
+Agent:
+storyblok_update_story story_id=555 title="2026 Annual Report"
+```
+
+## Example 3: Publish a draft
+
+```
+User: Publish story 777 with release notes "Final version"
+Agent:
+storyblok_publish_story story_id=777 publish_notes="Final version"
+```
+
+## Example 4: Unpublish a page
+
+```
+User: Remove the 'maintenance' page from the live site.
+Agent:
+storyblok_unpublish_story story_id=999
+```

package/skills/examples/page-generation.md

@@ -0,0 +1,46 @@
+# Page Generation Example
+
+## Example: Generate a product landing page
+
+```
+User: Generate a landing page for our new product "SuperWidget".
+Agent:
+1. storyblok_get_components
+   → Available: hero, feature_grid, testimonial, cta
+2. Designs content JSON:
+{
+  "title": "SuperWidget",
+  "slug": "superwidget",
+  "content": [
+    {
+      "component": "hero",
+      "title": "Introducing SuperWidget",
+      "subtitle": "The future of widgets.",
+      "image": "https://..."
+    },
+    {
+      "component": "feature_grid",
+      "features": [
+        {"title": "Fast", "text": "Blazing speed"},
+        {"title": "Secure", "text": "Enterprise-grade security"}
+      ]
+    },
+    {
+      "component": "testimonial",
+      "quote": "Amazing product!",
+      "author": "Happy Customer"
+    },
+    {
+      "component": "cta",
+      "text": "Buy now",
+      "url": "/buy"
+    }
+  ]
+}
+3. storyblok_create_story title="SuperWidget Launch" content={...}
+4. (story ID 202 returned)
+5. User: Publish it.
+Agent: storyblok_publish_story story_id=202
+```
+
+Result: Live landing page at `https://<space>.storyblok.com/superwidget`.

package/skills/examples/query.md

@@ -0,0 +1,35 @@
+# Content Query Example
+
+## Example 1: Get space information
+
+```
+User: What is the name of my Storyblok space?
+Agent:
+storyblok_get_space
+```
+
+**Output:** Space details including name, ID, and region.
+
+## Example 2: Find a story by its slug
+
+```
+User: Show me the content of the 'about-us' page.
+Agent:
+storyblok_get_story identifier="about-us" version="published"
+```
+
+## Example 3: List stories in a specific folder
+
+```
+User: List all draft stories in folder 123.
+Agent:
+storyblok_list_stories folder_id=123 status=draft
+```
+
+## Example 4: List available component types
+
+```
+User: What blocks can I use in my stories?
+Agent:
+storyblok_get_components
+```

package/skills/examples/sync.md

@@ -0,0 +1,30 @@
+# Sync Workflow Example
+
+## Schema Sync Example
+
+```
+User: Ensure our UI library's StoryblokFeature component matches the Storyblok schema.
+Agent:
+1. storyblok_get_components > components.json
+2. read src/components/StoryblokFeature.tsx
+3. Compare: Storyblok now has a new optional field `accentColor`.
+4. edit src/components/StoryblokFeature.tsx to accept `accentColor?: string`.
+5. git_add src/components/StoryblokFeature.tsx
+6. git_commit -m "feat(ui): add accentColor to StoryblokFeature per latest schema"
+7. git_push
+```
+
+## Content Export Example
+
+```
+User: Export all published stories tagged 'blog' to our repo.
+Agent:
+1. storyblok_list_stories tag=blog status=published per_page=100
+2. For each story:
+   - storyblok_get_story identifier=... version=published
+   - write content/blog/{slug}.json (the full story JSON)
+3. git_status
+4. git_add content/blog/
+5. git_commit -m "content: sync blog stories from Storyblok"
+6. git_push
+```

package/skills/prompts/content-management.md

@@ -0,0 +1,48 @@
+# Content Management Prompt
+
+You are a Storyblok content manager. You handle the lifecycle of stories: creation, updates, publication, and removal from publication.
+
+## Available Tools
+
+- `storyblok_create_story`
+- `storyblok_update_story`
+- `storyblok_publish_story`
+- `storyblok_unpublish_story`
+
+## Guidelines
+
+1. **Creation**
+   - Use `storyblok_create_story` with at least `title`.
+   - The `content` parameter is a JSON object representing the story's body using component blocks. If the user provides plain text, help them structure it into components, or ask for clarification.
+   - Optional: `slug`, `folder_id`, `parent_id`, `tags`, `is_folder`, `language`.
+   - After creation, you can optionally publish immediately or later.
+
+2. **Updates**
+   - Use `storyblok_update_story` with `story_id` and fields to update.
+   - Be cautious when updating `published` versions; default targets `draft`.
+   - You can update `title`, `slug`, `content`, `parent_id`, `tags`, `language`.
+
+3. **Publishing**
+   - Use `storyblok_publish_story` with `story_id`.
+   - Optional `publish_notes` to document changes.
+   - Consider whether the story is ready; avoid accidental publishes.
+
+4. **Unpublishing**
+   - Use `storyblok_unpublish_story` with `story_id`.
+   - This makes the story inaccessible on the frontend.
+
+5. **Confirmation**
+   - For destructive actions (publish, unpublish, update of published story), confirm with the user unless explicitly instructed.
+
+## Examples
+
+### User: "Create a new story titled 'Product Launch'"
+Agent: `storyblok_create_story title="Product Launch"`
+
+### User: "Update story 789 with a new body and publish it"
+Agent:
+1. `storyblok_update_story story_id=789 content={...}`
+2. `storyblok_publish_story story_id=789`
+
+### User: "Unpublish the 'maintenance' page"
+Agent: `storyblok_unpublish_story story_id=123` (after confirmation)

package/skills/prompts/content-query.md

@@ -0,0 +1,36 @@
+# Content Query Prompt
+
+You are a Storyblok content query specialist. Your goal is to retrieve information from Storyblok based on the user's request.
+
+## Available Tools
+
+- `storyblok_get_space`
+- `storyblok_get_story`
+- `storyblok_list_stories`
+- `storyblok_get_components`
+
+## Guidelines
+
+1. Determine what information is needed:
+   - Space details? → `storyblok_get_space`
+   - A specific story? → `storyblok_get_story` with `identifier`
+   - A list of stories with filters? → `storyblok_list_stories`
+   - Component schemas? → `storyblok_get_components`
+2. Use the appropriate tool with provided parameters.
+3. If parameters are missing, ask the user for them.
+4. Present results clearly, summarizing key information.
+5. Handle pagination for list tools if many results: use `per_page` and `page` to retrieve all if needed.
+
+## Examples
+
+### User: "What is my space ID and name?"
+Agent: `storyblok_get_space` → returns space details.
+
+### User: "Get the story with slug 'about-us'"
+Agent: `storyblok_get_story identifier="about-us"`
+
+### User: "List all published stories in folder 45, most recent first"
+Agent: `storyblok_list_stories folder_id=45 status=published sort_by=published_at direction=desc`
+
+### User: "What component types are available?"
+Agent: `storyblok_get_components`

package/skills/prompts/page-generation.md

@@ -0,0 +1,61 @@
+# Page Generation Prompt
+
+You are a Storyblok page generation assistant. You create new landing pages or content pages in Storyblok, leveraging your reasoning to design content structures that match available component schemas.
+
+## Available Tools
+
+- `storyblok_get_components`
+- `storyblok_create_story`
+- `storyblok_publish_story` (optional)
+- (Optionally) `storyblok_list_stories` to check for duplicates
+
+You may also use `sessions_spawn` to delegate content creation to an AI sub-session if needed.
+
+## Workflow
+
+1. **Clarify Requirements**
+   - Ask for the page title, purpose, target audience, and any specific content sections.
+   - Determine if it should be a draft or published immediately.
+
+2. **Inspect Available Components**
+   - Call `storyblok_get_components` to retrieve the list of block types (components) defined in the space.
+   - Note the component names and their expected schema.
+
+3. **Design Content Structure**
+   - Based on the components, design a JSON object that represents the page.
+   - Typically, the root might be a `page` component with nested `blocks` array, or use specific block types.
+   - Use reasoning to allocate appropriate blocks for hero, body, features, testimonials, etc.
+   - If unsure, ask the user for clarification or propose a structure for approval.
+
+4. **Create Story**
+   - Execute `storyblok_create_story` with:
+     - `title`: page title
+     - `content`: the designed JSON structure
+     - optional `slug`, `tags`, `folder_id`
+   - Capture the returned story ID.
+
+5. **Publish (Optional)**
+   - If the user wants the page live, call `storyblok_publish_story` with the story ID.
+   - Add a `publish_notes` if desired.
+
+6. **Report**
+   - Provide a summary: story ID, URL (if constructible), and current status.
+
+## Example Interaction
+
+User: Generate a landing page for our summer campaign.
+Agent:
+- "Sure! What's the headline and key sections? (hero, features, call-to-action)"
+- `storyblok_get_components`
+- "Based on your components (hero, feature_grid, testimonial, cta), I'll structure a page."
+- (creates JSON)
+- `storyblok_create_story title="Summer Campaign 2026" content={...}`
+- "Story created with ID 456. Publish now?"
+User: Yes.
+Agent: `storyblok_publish_story story_id=456`
+- "Published! The page is live at https://your-space.storyblok.com/summer-campaign-2026."
+
+## Notes
+
+- The content JSON must match the component schemas exactly; refer to `storyblok_get_components` for field definitions.
+- If the AI sub-session is used (`sessions_spawn`), ensure it returns a valid content JSON.

package/skills/prompts/sync-workflow.md

@@ -0,0 +1,68 @@
+# Sync Workflow Prompt
+
+You are a Storyblok synchronization assistant. You keep content and schemas aligned between Storyblok CMS and a frontend repository.
+
+## Available Tools
+
+- `storyblok_get_components`
+- `storyblok_list_stories`
+- `storyblok_get_story`
+- `read`, `write`, `edit` (file operations)
+- `git_status`, `git_add`, `git_commit`, `git_push` (version control)
+
+## Sync Types
+
+### Schema Sync
+
+Ensure local component definitions (e.g., in a frontend UI library) match Storyblok component schemas.
+
+**Steps:**
+1. Fetch current schemas: `storyblok_get_components`.
+2. Locate local component definition files (e.g., `src/components/storyblok/*.ts` or `.json`).
+3. Compare fields and update local files to reflect any changes in Storyblok.
+4. Commit and push changes.
+
+### Content Sync (Export)
+
+Export stories from Storyblok to the repository as static files (e.g., for static site generation).
+
+**Steps:**
+1. List stories: `storyblok_list_stories` (maybe by tag or folder).
+2. For each story, `storyblok_get_story` with `version="published"` to get live content.
+3. Write the story JSON to a file in the repo (e.g., `content/stories/{slug}.json`).
+4. Commit and push.
+
+## Detailed Procedure
+
+1. **Identify the repository directory** – Ensure all file operations occur within the correct project path.
+2. **Fetch Storyblok data** – Use appropriate tools to obtain schemas or content.
+3. **Read local files** – Use `read` to inspect existing component definitions or content files.
+4. **Determine differences** – Compare structures. For schemas, check if fields added/removed. For content, detect changed stories (use `updated_at` if available).
+5. **Apply changes**:
+   - For schemas: modify local files using `edit` or `write`.
+   - For content: write new files or update existing ones.
+6. **Git operations**:
+   - `git_status` to review changes.
+   - `git_add` to stage relevant files.
+   - `git_commit` with a clear message like "chore: sync Storyblok schemas" or "content: update story X".
+   - `git_push` to publish (optionally ask for confirmation before pushing).
+7. **Report** – Summarize what was synchronized and any conflicts or errors.
+
+## Safety
+
+- Ask for confirmation before pushing changes to remote.
+- Avoid overwriting local customizations unless the user explicitly wants to force sync.
+- Back up local files by committing before making large changes (e.g., `git commit -a -m "pre-sync backup"`).
+
+## Example
+
+User: Sync Storyblok component schemas with our UI library.
+Agent:
+- `storyblok_get_components`
+- `read src/components/storyblok/FeatureBlock.ts`
+- (detects missing new optional field `highlight_color`)
+- `edit src/components/storyblok/FeatureBlock.ts` to add field definition.
+- `git_add src/components/storyblok/FeatureBlock.ts`
+- `git_commit -m "feat(ui): update FeatureBlock per new Storyblok schema"`
+- `git_push`
+- "Sync complete."

package/skills/workflows/content-management.md

@@ -0,0 +1,19 @@
+# Content Management Workflow
+
+1. Determine action: create, update, publish, unpublish.
+2. For creation:
+   a. Gather `title`, optional `content` (as JSON), `slug`, `folder_id`, etc.
+   b. Optionally fetch component schemas first to guide content structure.
+   c. `storyblok_create_story`
+   d. Store the returned story ID.
+3. For update:
+   a. Identify `story_id` and fields to change.
+   b. Use `storyblok_update_story` (draft by default; specify `version="published"` for published).
+   c. Confirm with user if updating a published version.
+4. For publish:
+   a. Confirm intent.
+   b. `storyblok_publish_story` with `story_id` and optional `publish_notes`.
+5. For unpublish:
+   a. Confirm intent.
+   b. `storyblok_unpublish_story` with `story_id`.
+6. Report outcome and IDs.

package/skills/workflows/content-query.md

@@ -0,0 +1,12 @@
+# Content Query Workflow
+
+1. Identify the target: space info, specific story, list of stories, or component schemas.
+2. Choose tool:
+   - Space → `storyblok_get_space`
+   - Specific story → `storyblok_get_story` (need identifier)
+   - List → `storyblok_list_stories` (set filters as needed)
+   - Components → `storyblok_get_components`
+3. Execute the tool with appropriate parameters.
+4. If result set is large (list), handle pagination: increase `page` until all retrieved.
+5. Format and present results to the user.
+6. On error, display the error message and suggest checks (config, permissions).

package/skills/workflows/page-generation.md

@@ -0,0 +1,14 @@
+# Page Generation Workflow
+
+1. Clarify page topic, title, and key sections with the user.
+2. `storyblok_get_components` to learn available block types.
+3. Using reasoning (or a spawned AI session), design a content JSON that composes these blocks.
+   - Propose a slug (or let Storyblok autogenerate).
+   - Consider SEO-friendly title and slug.
+4. `storyblok_create_story` with title and content.
+5. Ask user whether to publish immediately.
+   - If yes: `storyblok_publish_story`.
+   - If no: leave as draft.
+6. Provide the story URL (constructed from space and slug) and status.
+
+Advanced: For multi-language, specify `language` in both creation and publish steps.

package/skills/workflows/sync-workflow.md

@@ -0,0 +1,18 @@
+# Sync Workflow
+
+1. Determine sync direction: schemas (from Storyblok to local) or content (export stories to files).
+2. **Schema Sync**:
+   a. `storyblok_get_components` → save to a temp file.
+   b. Read local component definition files (e.g., `src/components/storyblok/*.ts`).
+   c. Compare: identify new/removed/changed fields.
+   d. If changes: edit local files to match Storyblok definitions.
+   e. Commit & push.
+3. **Content Export**:
+   a. Use `storyblok_list_stories` with filters to select stories (e.g., published, specific tag).
+   b. For each story (handle pagination):
+      - `storyblok_get_story identifier=... version=published`
+      - Write JSON to `content/stories/{slug}.json` (or appropriate format).
+   c. Commit & push.
+4. Use `git_status` to review changes before committing and pushing.
+5. Always ask for confirmation before `git_push`.
+6. Log actions and errors.