docs(plugin.xml): refresh marketplace description with current features

The marketplace description was stale: it listed 7 LLM providers (vs.
17 actually supported), omitted Agent Mode entirely, and bundled
"Custom Skills: User-defined prompts and commands" — confusing now
that Skills (LLM-activated SKILL.md) and Commands (slash commands)
are distinct things in v1.5.0.

Rewrote Key Features to cover Agent Mode, Skills, Commands, SDD,
Security Scanning, Inline Completion, ACP/CLI Runners, and the
Plugin Integration API, and expanded the LLM list to match what the
README already advertises. Updated Quick Start to mention Llama.cpp,
Exo, and the Skills directories.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

Author: stephanj

src/main/resources/META-INF/plugin.xml

@@ -23,21 +23,28 @@
 
         <h3>Key Features</h3>
         <UL>
-            <LI><strong>Local & Cloud LLMs:</strong> Ollama, LMStudio, GPT4All, OpenAI, Anthropic, Gemini, Groq, and more</LI>
-            <LI><strong>MCP Support:</strong> Model Context Protocol with integrated marketplace for installing servers</LI>
-            <LI><strong>RAG:</strong> Retrieval-Augmented Generation for semantic code search</LI>
+            <LI><strong>Local &amp; Cloud LLMs:</strong> Ollama, LMStudio, GPT4All, Llama.cpp, Exo + OpenAI, Anthropic, Gemini, Mistral, Groq, DeepInfra, DeepSeek, Kimi, GLM, OpenRouter, Azure OpenAI, Amazon Bedrock</LI>
+            <LI><strong>Agent Mode:</strong> Autonomous code tools with parallel sub-agents, multi-turn tool use, and built-in file/search/run_command tools</LI>
+            <LI><strong>Skills (new in v1.5.0):</strong> Portable <code>SKILL.md</code> files the LLM activates mid-conversation — the same files also work in Claude Code, Codex and Gemini</LI>
+            <LI><strong>Commands:</strong> User-defined slash commands (previously "Custom Prompts") for reusable prompt templates</LI>
+            <LI><strong>MCP Support:</strong> Model Context Protocol with integrated marketplace for installing servers (stdio and HTTP/SSE transports)</LI>
+            <LI><strong>Spec Driven Development:</strong> Backlog.md task management with Spec Browser, Kanban Board, and Agent Loop batch execution with dependency ordering</LI>
+            <LI><strong>Security Scanning:</strong> Run Gitleaks (secrets), OpenGrep (SAST) and Trivy (CVEs) as agent tools — findings auto-create prioritised backlog tasks</LI>
+            <LI><strong>Inline Code Completion:</strong> AI-powered suggestions as you type using Fill-in-the-Middle models via Ollama or LM Studio</LI>
+            <LI><strong>ACP &amp; CLI Runners:</strong> Integrate external agents (Claude Code, Codex, Gemini CLI, Kimi, Copilot) via the Agent Communication Protocol or CLI invocation</LI>
+            <LI><strong>RAG:</strong> Retrieval-Augmented Generation for semantic code search (ChromaDB + Ollama embeddings)</LI>
             <LI><strong>Web Search:</strong> Integrate Google and Tavily search into your prompts</LI>
-            <LI><strong>Custom Skills:</strong> User-defined prompts and commands</LI>
-            <LI><strong>Chat History:</strong> Locally stored conversations</LI>
-            <LI><strong>Token Cost Calculator:</strong> Real-time cost estimation for cloud providers</LI>
-            <LI><strong>Multi-modal Support:</strong> Drag & drop images for vision models</LI>
+            <LI><strong>Plugin Integration API:</strong> Other IntelliJ plugins can send findings to DevoxxGenie at runtime via a reflection-based <code>ExternalPromptService</code></LI>
+            <LI><strong>Chat History:</strong> Locally stored conversations (SQLite) with parallel conversation tabs</LI>
+            <LI><strong>Token Cost Calculator:</strong> Real-time cost and context-window estimation for cloud providers</LI>
+            <LI><strong>Multi-modal Support:</strong> Drag &amp; drop images for vision models, plus Git diff and project-context attachments</LI>
         </UL>
         <h3>Quick Start</h3>
         <UL>
             <LI>Install the plugin</LI>
-            <LI>For local LLMs: Install Ollama, LMStudio, or GPT4All</LI>
-            <LI>For cloud LLMs: Add your API keys in Settings</LI>
-            <LI>Optional: Configure MCP servers from the marketplace</LI>
+            <LI>For local LLMs: install Ollama, LMStudio, GPT4All, Llama.cpp or Exo</LI>
+            <LI>For cloud LLMs: add your API keys in <em>Settings → DevoxxGenie → LLM Providers</em></LI>
+            <LI>Optional: enable Agent Mode, configure MCP servers from the marketplace, or drop a <code>SKILL.md</code> into <code>.devoxxgenie/skills/</code>, <code>.claude/skills/</code> or <code>.agents/skills/</code></LI>
             <LI>Start prompting!</LI>
         </UL>