ima-claude 2.9.0

package/plugins/ima-claude/skills/livecanvas/references/picostrap.md

@@ -0,0 +1,227 @@
+# PicoStrap Theme Integration
+
+PicoStrap5 — free, open-source Bootstrap 5 starter theme by the LiveCanvas team. Ultra-lightweight, no bloat.
+
+**GitHub**: https://github.com/livecanvas-team/picostrap5
+
+## Table of Contents
+
+1. [Overview](#overview)
+2. [Architecture](#architecture)
+3. [SCSS Pipeline](#scss-pipeline)
+4. [Child Theme Setup](#child-theme-setup)
+5. [NinjaBootstrap](#ninjabootstrap)
+6. [Dark Mode](#dark-mode)
+7. [Customizer Options](#customizer-options)
+
+---
+
+## Overview
+
+- Pure Bootstrap 5.3 — no extras, no jQuery
+- Built-in browser-based SCSS compiler (no CLI tools needed)
+- Hit Publish in Customizer → Bootstrap recompiles on the fly
+- jQuery-free (vanilla JS Bootstrap)
+- No icon fonts (inline SVGs)
+- RFS (Responsive Font Sizing) enabled by default
+- WooCommerce support out of the box
+- Clean HTML head (removes WP default header bloat)
+- GPL licensed
+- LiveCanvas auto-detects PicoStrap5 and enables Bootstrap 5 mode
+
+## Architecture
+
+```
+picostrap5/
+├── header.php
+├── footer.php
+├── index.php
+├── page.php
+├── single.php
+├── archive.php
+├── search.php
+├── 404.php
+├── functions.php
+├── inc/
+│   ├── template-tags.php        # Hooks and filters
+│   └── clean-head.php           # Remove WP bloat
+├── sass/
+│   ├── main.scss                # Entry point
+│   ├── _bootstrap-loader.scss   # Bootstrap import
+│   └── _custom.scss             # Theme customizations
+├── css-output/
+│   └── main.css                 # Compiled output
+├── loops/                       # Loop templates
+├── partials/                    # Template parts
+├── woocommerce/                 # WC template overrides
+└── livecanvas/
+    ├── pages/                   # Readymade page templates
+    └── sections/                # Custom section library
+```
+
+## SCSS Pipeline
+
+Compilation order:
+
+```
+main.scss
+  └── _bootstrap-loader.scss
+        ├── Bootstrap 5.3 source SCSS
+        ├── NinjaBootstrap extensions
+        └── _custom.scss (your overrides)
+```
+
+### Variable Override Strategy
+
+Bootstrap uses `!default` — your variables take precedence when defined before Bootstrap's imports:
+
+```scss
+// In child theme sass/_custom.scss
+$primary: #your-brand-color;
+$font-family-base: 'Your Font', sans-serif;
+$border-radius: 0.5rem;
+
+// These override Bootstrap defaults because they're loaded first
+```
+
+### Where to Customize
+
+| What | Where |
+|------|-------|
+| Bootstrap variable overrides | Child theme `sass/_custom.scss` |
+| Custom component styles | Child theme `sass/_custom.scss` |
+| Theme structure/templates | Child theme PHP files |
+| LiveCanvas page templates | Child theme `livecanvas/pages/` |
+| LiveCanvas sections | Child theme `livecanvas/sections/` |
+
+### Recompiling
+
+1. Go to Appearance → Customize
+2. Change any SASS variable or save `_custom.scss`
+3. Hit Publish — SCSS recompiles in browser
+4. Output → `css-output/main.css`
+
+No Node.js, no webpack, no CLI tools required.
+
+---
+
+## Child Theme Setup
+
+PicoStrap provides a blank child theme starter:
+
+```
+picostrap5-child/
+├── style.css                    # Theme header (Template: picostrap5)
+├── functions.php                # Child functions
+├── sass/
+│   └── _custom.scss             # Bootstrap variable overrides + custom CSS
+├── livecanvas/
+│   ├── pages/                   # Custom readymade page templates
+│   └── sections/                # Custom section library
+└── woocommerce/                 # WC template overrides (optional)
+```
+
+### Key Override Points
+
+**Colors**:
+```scss
+$primary:   #your-color;
+$secondary: #your-color;
+$success:   #your-color;
+$info:      #your-color;
+$warning:   #your-color;
+$danger:    #your-color;
+$light:     #your-color;
+$dark:      #your-color;
+```
+
+**Typography**:
+```scss
+$font-family-base: 'Your Font', system-ui, sans-serif;
+$font-family-monospace: 'Your Mono', monospace;
+$font-size-base: 1rem;
+$line-height-base: 1.5;
+$headings-font-family: 'Your Heading Font', serif;
+$headings-font-weight: 700;
+```
+
+**Spacing & Sizing**:
+```scss
+$spacer: 1rem;
+$border-radius: 0.375rem;
+$border-radius-lg: 0.5rem;
+$border-radius-sm: 0.25rem;
+```
+
+**Components**:
+```scss
+$btn-border-radius: 0.375rem;
+$card-border-radius: 0.5rem;
+$input-border-radius: 0.375rem;
+$navbar-padding-y: 0.5rem;
+```
+
+For IMA-specific overrides, see the `ima-bootstrap` skill's `references/theme-integration.md`.
+
+---
+
+## NinjaBootstrap
+
+Built into PicoStrap v3+. Extends Bootstrap 5.3 with additional utility classes:
+
+- **10 color shades per theme color**: `.text-primary-100` through `.text-primary-900`, `.bg-primary-100` through `.bg-primary-900`
+- **Automatic dark mode inversion**: Shades flip appropriately in `[data-bs-theme="dark"]`
+- **Extra utilities**: Additional spacing, sizing, and display helpers beyond Bootstrap defaults
+
+Usage in LiveCanvas:
+```html
+<section class="bg-primary-100 py-5">
+  <div class="container">
+    <h2 class="text-primary-800">Section Title</h2>
+    <p class="text-primary-600">Subtitle text with lighter shade</p>
+  </div>
+</section>
+```
+
+---
+
+## Dark Mode
+
+PicoStrap supports Bootstrap 5.3's native dark mode via `data-bs-theme` attribute:
+
+```html
+<!-- Page-level -->
+<html data-bs-theme="dark">
+
+<!-- Component-level -->
+<div data-bs-theme="dark" class="card">
+  <div class="card-body">Always dark</div>
+</div>
+
+<!-- Toggle via JavaScript -->
+<script>
+document.documentElement.setAttribute('data-bs-theme',
+  document.documentElement.getAttribute('data-bs-theme') === 'dark' ? 'light' : 'dark'
+);
+</script>
+```
+
+NinjaBootstrap color shades automatically invert in dark mode.
+
+---
+
+## Customizer Options
+
+PicoStrap exposes key Bootstrap variables through the WordPress Customizer:
+
+- **Colors**: All theme colors (primary, secondary, etc.)
+- **Typography**: Font families, sizes, weights
+- **Spacing**: Base spacer value
+- **Components**: Border radius, button styles
+- **AI Color Palette**: Generate harmonious color schemes
+
+Changes trigger SCSS recompilation on Publish.
+
+### Export/Import
+
+PicoStrap supports Customizer Export/Import for sharing design system settings across sites. Useful for team standardization.

package/plugins/ima-claude/skills/mcp-atlassian/SKILL.md

@@ -0,0 +1,339 @@
+---
+name: mcp-atlassian
+description: >-
+  Atlassian MCP for Jira and Confluence operations (Claude's bundled integration).
+  Token-efficient workflows for issue management, page creation, search, and user mentions.
+  Use when: creating/editing Jira issues, searching with JQL/CQL, creating/updating Confluence
+  pages, adding comments, transitioning issue status, looking up users, mentioning/tagging
+  users in Jira or Confluence content, or any Atlassian Cloud operation. Triggers on: Jira,
+  Confluence, JQL, CQL, sprint, epic, story, issue, wiki page, Atlassian, @mention in
+  Jira/Confluence context. Provides 50-80% token savings over naive API usage through
+  field filtering, pagination control, and format selection.
+---
+
+# Atlassian MCP - Jira & Confluence Integration
+
+Claude's bundled Atlassian MCP. All tools prefixed `mcp__claude_ai_Atlassian__`.
+
+## Bootstrap (Required First Call)
+
+Every session MUST start with `getAccessibleAtlassianResources` to obtain the `cloudId`.
+Nearly all other tools require it. Cache this value mentally for the session.
+
+```
+getAccessibleAtlassianResources  → returns cloudId (UUID or site URL)
+```
+
+## Tool Catalog
+
+### Discovery & Users
+
+| Tool | Purpose | Key Params |
+|------|---------|------------|
+| `getAccessibleAtlassianResources` | Get cloudId (CALL FIRST) | *(none)* |
+| `atlassianUserInfo` | Current authenticated user | *(none)* |
+| `lookupJiraAccountId` | Find user accountId by name/email | `cloudId`, `searchString` |
+
+### Search (3 tools - choose wisely)
+
+| Tool | Query Language | Scope | When to Use |
+|------|---------------|-------|-------------|
+| `search` | Natural language | Jira + Confluence | Default search. Always use unless JQL/CQL specifically needed |
+| `searchJiraIssuesUsingJql` | JQL | Jira only | Structured queries: status, assignee, project, date ranges |
+| `searchConfluenceUsingCql` | CQL | Confluence only | Structured queries: space, label, type, creator |
+
+### Jira - Read
+
+| Tool | Purpose | Token-Saving Tip |
+|------|---------|------------------|
+| `getJiraIssue` | Get issue details | Use `fields` param to request ONLY needed fields |
+| `getVisibleJiraProjects` | List projects | Use `searchString` to filter |
+| `getJiraProjectIssueTypesMetadata` | Issue types for a project | Call before `createJiraIssue` |
+| `getJiraIssueTypeMetaWithFields` | Field metadata for issue type | Call before `editJiraIssue` to know valid fields |
+| `getTransitionsForJiraIssue` | Available status transitions | MUST call before `transitionJiraIssue` |
+| `getJiraIssueRemoteIssueLinks` | Remote links on an issue | |
+
+### Jira - Write
+
+| Tool | Content Format | Purpose |
+|------|---------------|---------|
+| `createJiraIssue` | **Markdown** (description) | Create issue. Has `assignee_account_id` param |
+| `editJiraIssue` | **Raw fields object** | Update any fields. Use for ADF descriptions with mentions |
+| `addCommentToJiraIssue` | **Markdown** | Add comment. Has `commentVisibility` for restricted comments |
+| `transitionJiraIssue` | Transition object | Change status. Get transition IDs first |
+| `addWorklogToJiraIssue` | Duration string | Log time: `"2h"`, `"30m"`, `"4d"` |
+
+### Confluence - Read
+
+| Tool | Purpose | Token-Saving Tip |
+|------|---------|------------------|
+| `getConfluencePage` | Get page by ID | Use `contentFormat: "markdown"` for smaller responses |
+| `getConfluenceSpaces` | List spaces | Filter with `keys`, `type`, `labels` |
+| `getPagesInConfluenceSpace` | Pages in a space | Filter with `title`, `status`, `sort` |
+| `getConfluencePageDescendants` | Child pages | Use `depth` to limit |
+| `getConfluencePageFooterComments` | Footer comments | |
+| `getConfluencePageInlineComments` | Inline comments | Filter by `resolutionStatus` |
+
+### Confluence - Write
+
+| Tool | Content Format | Purpose |
+|------|---------------|---------|
+| `createConfluencePage` | **Markdown** or **ADF** | Create page. Requires `spaceId` (not space key!) |
+| `updateConfluencePage` | **Markdown** or **ADF** | Replace entire page body |
+| `createConfluenceFooterComment` | **Markdown only** | Comment on page |
+| `createConfluenceInlineComment` | **Markdown only** | Comment on specific text selection |
+
+### Cross-Product
+
+| Tool | Purpose |
+|------|---------|
+| `search` | Rovo natural language search across Jira + Confluence |
+| `fetch` | Get detail by ARI (Atlassian Resource Identifier). Read-only |
+
+## User Mentions (@tagging)
+
+**This is the most error-prone area.** Follow these patterns exactly.
+
+### Step 1: Always Look Up the accountId First
+
+```
+lookupJiraAccountId
+  cloudId: "<your-cloud-id>"
+  searchString: "john@example.com"   # or "John Doe" or partial "john"
+```
+
+Returns users with `accountId` values. This works for BOTH Jira and Confluence
+(same Atlassian org user pool).
+
+### Step 2: Choose the Right Mention Strategy
+
+#### Confluence Pages (RELIABLE - use ADF)
+
+Use `createConfluencePage` or `updateConfluencePage` with `contentFormat: "adf"`:
+
+```json
+{
+  "cloudId": "<cloud-id>",
+  "spaceId": "<space-id>",
+  "title": "Page Title",
+  "contentFormat": "adf",
+  "body": "{\"version\":1,\"type\":\"doc\",\"content\":[{\"type\":\"paragraph\",\"content\":[{\"type\":\"text\",\"text\":\"Assigned to \"},{\"type\":\"mention\",\"attrs\":{\"id\":\"5b10a2844c20165700ede21g\",\"text\":\"@John Doe\"}},{\"type\":\"text\",\"text\":\" for review.\"}]}]}"
+}
+```
+
+**Critical:** The `body` value MUST be a JSON **string** (result of JSON.stringify),
+NOT a raw object. This is the #1 cause of failures.
+
+ADF mention node structure:
+```json
+{
+  "type": "mention",
+  "attrs": {
+    "id": "<accountId>",
+    "text": "@Display Name"
+  }
+}
+```
+
+The mention node is **inline** - it MUST be inside a `paragraph.content` array,
+never at the top-level `doc.content`.
+
+#### Confluence Comments (Markdown only - no native mentions)
+
+`createConfluenceFooterComment` and `createConfluenceInlineComment` accept only
+Markdown. Real @mentions are not supported in Markdown mode. Workaround:
+reference the user by name in plain text, or create/update the parent **page**
+with ADF to include the mention there instead.
+
+#### Jira Issue Description (RELIABLE - use editJiraIssue with ADF)
+
+For mentions in descriptions, use `editJiraIssue` with the ADF description field:
+
+```json
+{
+  "cloudId": "<cloud-id>",
+  "issueIdOrKey": "PROJ-123",
+  "fields": {
+    "description": {
+      "version": 1,
+      "type": "doc",
+      "content": [
+        {
+          "type": "paragraph",
+          "content": [
+            { "type": "text", "text": "Assigned to " },
+            {
+              "type": "mention",
+              "attrs": {
+                "id": "5b10a2844c20165700ede21g",
+                "text": "@John Doe"
+              }
+            },
+            { "type": "text", "text": " for review." }
+          ]
+        }
+      ]
+    }
+  }
+}
+```
+
+**Note:** `createJiraIssue` takes Markdown for description. To create an issue
+WITH mentions, create the issue first (Markdown description), then immediately
+`editJiraIssue` to replace the description with ADF containing mentions.
+
+#### Jira Comments (Markdown - limited mention support)
+
+`addCommentToJiraIssue` accepts Markdown via `commentBody`. Real ADF mentions
+are not directly available. Workaround: reference users by display name in the
+Markdown text. If real @mentions in comments are critical, use a two-step
+approach: add the comment, then use the Jira UI or a direct API call for the
+mention.
+
+### Mention Pitfalls Checklist
+
+| Pitfall | Symptom | Fix |
+|---------|---------|-----|
+| Using `username` instead of `accountId` | Shows "unlicensed user" | Always use `lookupJiraAccountId` first |
+| ADF body as object instead of string | "invalid request body" error | `JSON.stringify()` the ADF for Confluence |
+| Mention at doc level, not in paragraph | Silent failure / no mention | Wrap in `{"type": "paragraph", "content": [...]}` |
+| Missing `<ac:link>` wrapper (storage format) | Not rendered as mention | N/A for this MCP (uses ADF/Markdown, not storage format) |
+| Wiki notation `[~accountId:xxx]` in Markdown | Raw text, not a mention | Use ADF via `editJiraIssue` instead |
+| Using `@username` in Markdown | Plain text, not a real mention | Use ADF path for real mentions |
+
+## Token-Saving Strategies
+
+### 1. Filter Jira Fields
+
+```
+getJiraIssue
+  fields: ["summary", "status", "assignee", "priority"]  # NOT the full issue
+```
+
+Never fetch all fields. Common useful subsets:
+- **Quick status check:** `["summary", "status", "priority"]`
+- **Assignment info:** `["summary", "assignee", "reporter"]`
+- **Planning:** `["summary", "status", "priority", "issuetype", "parent"]`
+
+### 2. Use Markdown for Confluence Reads
+
+```
+getConfluencePage
+  contentFormat: "markdown"  # Much smaller than ADF
+```
+
+ADF responses can be 5-10x larger than Markdown equivalents.
+
+### 3. Limit Search Results
+
+```
+searchJiraIssuesUsingJql
+  maxResults: 10       # Default is 50, often excessive
+  fields: ["summary", "status"]  # Minimal fields
+```
+
+### 4. Use Rovo Search for Discovery, JQL/CQL for Precision
+
+- "Find issues about authentication" → `search` (natural language, cross-product)
+- "All open bugs in PROJ assigned to me" → `searchJiraIssuesUsingJql` with JQL
+- "Pages labeled 'architecture' in DEV space" → `searchConfluenceUsingCql` with CQL
+
+### 5. Avoid Redundant Discovery Calls
+
+Cache these within a session (they rarely change):
+- `cloudId` from `getAccessibleAtlassianResources`
+- Project keys from `getVisibleJiraProjects`
+- Space IDs from `getConfluenceSpaces`
+- Issue type IDs from `getJiraProjectIssueTypesMetadata`
+
+## Common Workflows
+
+### Create and Assign a Jira Issue
+
+```
+1. getAccessibleAtlassianResources → cloudId
+2. lookupJiraAccountId(searchString: "jane@co.com") → accountId
+3. createJiraIssue(
+     projectKey: "PROJ",
+     issueTypeName: "Task",
+     summary: "Review Q4 report",
+     description: "Review and approve the Q4 financial report",
+     assignee_account_id: "<accountId>"
+   )
+```
+
+### Transition a Jira Issue
+
+```
+1. getTransitionsForJiraIssue(issueIdOrKey: "PROJ-123") → find transition ID
+2. transitionJiraIssue(
+     issueIdOrKey: "PROJ-123",
+     transition: { "id": "<transitionId>" }
+   )
+```
+
+### Create a Confluence Page with User Mention
+
+```
+1. getAccessibleAtlassianResources → cloudId
+2. lookupJiraAccountId(searchString: "john") → accountId
+3. getConfluenceSpaces(keys: ["DEV"]) → spaceId
+4. createConfluencePage(
+     spaceId: "<spaceId>",
+     title: "Meeting Notes",
+     contentFormat: "adf",
+     body: "<ADF JSON string with mention node>"
+   )
+```
+
+### Inline Comment on Specific Text
+
+```
+createConfluenceInlineComment(
+  pageId: "<pageId>",
+  body: "This section needs updating",
+  inlineCommentProperties: {
+    "textSelection": "exact text on page",
+    "textSelectionMatchCount": 1,
+    "textSelectionMatchIndex": 0
+  }
+)
+```
+
+## Decision Logic
+
+```
+Need to search across Jira AND Confluence?
+  → search (Rovo, natural language)
+
+Need structured Jira query (by project/status/assignee/date)?
+  → searchJiraIssuesUsingJql (JQL)
+
+Need structured Confluence query (by space/label/type)?
+  → searchConfluenceUsingCql (CQL)
+
+Need to @mention a user in content?
+  → lookupJiraAccountId first, then:
+    - Confluence page → ADF with mention node
+    - Jira description → editJiraIssue with ADF fields
+    - Jira/Confluence comment → Plain text name (Markdown limitation)
+
+Need issue field metadata before editing?
+  → getJiraIssueTypeMetaWithFields
+
+Need to change issue status?
+  → getTransitionsForJiraIssue THEN transitionJiraIssue
+
+Creating a Confluence page?
+  → Need spaceId (NOT space key). Get from getConfluenceSpaces
+```
+
+## What This MCP Does NOT Support
+
+- Direct Atlassian storage format (XML) - uses Markdown or ADF only
+- Confluence page permissions management
+- Jira board/sprint management
+- Jira custom field creation
+- Attachment uploads
+- Bulk operations
+- Confluence page templates

package/plugins/ima-claude/skills/mcp-context7/SKILL.md

@@ -0,0 +1,109 @@
+---
+name: mcp-context7
+description: Use Context7 MCP for official library documentation, framework APIs, best practices, code examples. Triggers on library/framework questions like 'how to use [library]', 'show [library] example', '[library] docs', '[library] API', or framework names (React, Vue, Quasar, Next.js, Express, Fastify, Prisma, tRPC, Zod, Tailwind, etc). Also triggers on import statements detected in code or package.json library lookups. Provides 70-80% token savings over web searching.
+---
+
+# Context7 MCP - Library Documentation
+
+Use Context7 for official library documentation instead of web searching or guessing APIs.
+
+## MCP Tools
+
+| Tool | Purpose |
+|------|---------|
+| `mcp__context7__search` | Search for libraries and get documentation |
+
+## Basic Usage
+
+Context7 combines library resolution and documentation retrieval into a single tool.
+
+```
+mcp__context7__search
+  query: "How to use QDialog component in Quasar"
+```
+
+**Parameters**:
+- `query` (required): Your question or search query including the library name
+
+The tool will:
+1. Identify the library from your query
+2. Find the relevant documentation
+3. Return focused, relevant docs
+
+## Query Optimization
+
+**Be specific and include**:
+- Component/function names: "QDialog component API props events slots"
+- What you want to do: "How to set up authentication with JWT"
+- Context: "React useState hook example with TypeScript"
+
+**Good queries**:
+| Query | Why It's Good |
+|-------|---------------|
+| "Quasar QDialog props and events" | Specific component, clear intent |
+| "React useEffect cleanup function" | Specific hook, specific aspect |
+| "Prisma findMany where clause syntax" | Specific method, specific feature |
+| "Express middleware error handling" | Framework + feature |
+
+**Avoid vague queries**:
+- ❌ "How does Quasar work?"
+- ✅ "How to create a Quasar button with icon"
+
+## Decision Logic
+
+```
+IF question about library/framework API:
+    → Use Context7
+ELSE IF import statement detected AND user asks about that library:
+    → Use Context7
+ELSE IF general programming concept (closures, promises, etc.):
+    → Use native Claude knowledge
+ELSE IF library not found:
+    → Fallback to WebSearch or Tavily
+ELSE IF asking for "latest" or "new" features post-cutoff:
+    → Use Tavily instead
+```
+
+## When NOT to Use
+
+- General programming questions (no specific library)
+- Debugging code that doesn't involve library APIs
+- Simple syntax questions Claude already knows
+- User wants current/latest info post-cutoff (use Tavily instead)
+
+## Common Libraries Supported
+
+**Frontend**: React, Vue, Quasar, Next.js, Nuxt, Svelte, Angular, Tailwind, Bootstrap
+**Backend**: Express, Fastify, Nest.js, tRPC, Prisma, Sequelize, TypeORM
+**Utilities**: Lodash, Ramda, date-fns, Zod, Yup, Joi
+**Build**: Vite, Webpack, Rollup, ESBuild
+**Testing**: Jest, Vitest, Playwright, Cypress
+
+## Examples
+
+| User Request | Action |
+|--------------|--------|
+| "How to use QDialog in Vue?" | search(query: "QDialog component Quasar Vue") |
+| "React useState example" | search(query: "React useState hook example") |
+| "Prisma query syntax" | search(query: "Prisma findMany where query") |
+| "What's a closure?" | Native Claude (no library) |
+| "Latest React 19 features" | Use Tavily (current info needed) |
+
+## Multiple Queries
+
+If initial results aren't sufficient, refine your query:
+1. First attempt: "Quasar form validation"
+2. If needed: "Quasar QForm validation rules API"
+3. If needed: "Quasar field validation with Vuelidate"
+
+## Setup
+
+No API key required. Install with:
+```bash
+bun run scripts/setup-mcp.ts
+```
+
+Or manually:
+```bash
+claude mcp add --scope user context7 -- npx -y @upstash/context7-mcp@latest
+```