@rlabs-inc/gemini-mcp 0.5.0 → 0.5.1

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/README.md CHANGED
@@ -4,33 +4,45 @@ A Model Context Protocol (MCP) server for integrating Google's Gemini 3 models w
4
4
 
5
5
  [![npm version](https://badge.fury.io/js/@rlabs-inc%2Fgemini-mcp.svg)](https://www.npmjs.com/package/@rlabs-inc/gemini-mcp)
6
6
 
7
- ## What's New in v0.4.0
7
+ ## What's New in v0.5.1
8
8
 
9
- **20+ tools** for comprehensive Gemini 3 integration:
9
+ **30+ tools** for comprehensive Gemini 3 integration - the most complete Gemini MCP server available!
10
+
11
+ **Text-to-Speech (NEW!):**
12
+ - **gemini-speak** - Convert text to speech with 30 unique voices
13
+ - **gemini-dialogue** - Generate two-speaker conversations
14
+ - **gemini-list-voices** - Browse all available voices
15
+
16
+ **URL Analysis (NEW!):**
17
+ - **gemini-analyze-url** - Analyze web pages with questions
18
+ - **gemini-compare-urls** - Compare two URLs side by side
19
+ - **gemini-extract-from-url** - Extract structured data from pages
20
+
21
+ **Context Caching (NEW!):**
22
+ - **gemini-create-cache** - Cache large documents for repeated queries
23
+ - **gemini-query-cache** - Query cached content efficiently
24
+ - **gemini-list-caches** / **gemini-delete-cache** - Manage caches
10
25
 
11
26
  **Multimodal Analysis:**
12
27
  - **YouTube Analysis** - Analyze videos by URL with timestamps and clipping
13
28
  - **Document Analysis** - PDFs, DOCX, spreadsheets with table extraction
14
29
 
15
30
  **Generation & Editing:**
16
- - **4K Image Generation** - Up to 4K resolution with Nano Banana Pro
17
- - **10 Aspect Ratios** - 1:1, 2:3, 3:2, 3:4, 4:3, 4:5, 5:4, 9:16, 16:9, 21:9
31
+ - **4K Image Generation** - Up to 4K resolution with 10 aspect ratios
18
32
  - **Multi-Turn Image Editing** - Iteratively refine images through conversation
19
- - **Google Search Grounding** - Ground images in real-world information
33
+ - **Video Generation** - Create videos with Veo 2.0
20
34
 
21
35
  **Advanced Tools:**
22
- - **Code Execution** - Gemini writes AND runs Python code (pandas, matplotlib, numpy, etc.)
36
+ - **Code Execution** - Gemini writes AND runs Python code
23
37
  - **Google Search** - Real-time web information with citations
24
38
  - **Structured Output** - JSON schema responses with validation
25
- - **Data Extraction** - Extract entities, facts, sentiment from text
26
-
27
- **Core Improvements:**
28
- - **Thinking Levels** - Control reasoning depth: minimal, low, medium, high
29
- - **Gemini 3 Models** - Updated to latest frontier models
39
+ - **Brainstorming** - Claude + Gemini collaborative problem-solving
30
40
 
31
41
  ### Previous Versions
32
42
 
33
- **v0.3.0:** Phase 2-3 features (thinking levels, code execution, search)
43
+ **v0.5.0:** 30+ tools, TTS, URL analysis, caching
44
+ **v0.4.0:** YouTube, documents, code execution, search
45
+ **v0.3.0:** Thinking levels, structured output
34
46
  **v0.2.0:** Image/Video generation with Veo
35
47
 
36
48
  ---
@@ -39,18 +51,21 @@ A Model Context Protocol (MCP) server for integrating Google's Gemini 3 models w
39
51
 
40
52
  | Feature | Description |
41
53
  |-------------------------------|-----------------------------------------------------------------|
54
+ | **Text-to-Speech** | 30 unique voices, single speaker or two-speaker dialogues |
55
+ | **URL Analysis** | Analyze, compare, and extract data from web pages |
56
+ | **Context Caching** | Cache large documents for efficient repeated queries |
42
57
  | **YouTube Analysis** | Analyze videos by URL with timestamp clipping |
43
58
  | **Document Analysis** | PDFs, DOCX, spreadsheets with table extraction |
44
59
  | **4K Image Generation** | Generate images up to 4K with 10 aspect ratios |
45
60
  | **Multi-Turn Image Editing** | Iteratively refine images through conversation |
46
- | **Video Generation** | Create videos with Veo (async with polling) |
61
+ | **Video Generation** | Create videos with Veo 2.0 (async with polling) |
47
62
  | **Code Execution** | Gemini writes and runs Python code (pandas, numpy, matplotlib) |
48
63
  | **Google Search** | Real-time web information with inline citations |
49
64
  | **Structured Output** | JSON responses with schema validation |
50
65
  | **Data Extraction** | Extract entities, facts, sentiment from text |
51
66
  | **Thinking Levels** | Control reasoning depth (minimal/low/medium/high) |
52
67
  | **Direct Query** | Send prompts to Gemini 3 Pro/Flash models |
53
- | **Brainstorming** | Collaborative problem-solving |
68
+ | **Brainstorming** | Claude + Gemini collaborative problem-solving |
54
69
  | **Code Analysis** | Analyze code for quality, security, performance |
55
70
  | **Summarization** | Summarize content at different detail levels |
56
71
 
@@ -6,6 +6,8 @@
6
6
  import { z } from "zod";
7
7
  import { generateWithGeminiPro } from "../gemini-client.js";
8
8
  import { logger } from "../utils/logger.js";
9
+ /** Consensus threshold - score at which brainstorming is considered complete */
10
+ const CONSENSUS_THRESHOLD = 8;
9
11
  /**
10
12
  * Register brainstorm tool with the MCP server
11
13
  */
@@ -58,7 +60,7 @@ Format this as: "Consensus Score: [NUMBER]"
58
60
  consensusScore: consensusScore
59
61
  });
60
62
  // Check if we already have consensus
61
- if (consensusScore >= 8) {
63
+ if (consensusScore >= CONSENSUS_THRESHOLD) {
62
64
  logger.info(`Consensus reached in first round with score ${consensusScore}`);
63
65
  consensusReached = true;
64
66
  }
@@ -143,7 +145,7 @@ Format: "Consensus Score: [NUMBER]"
143
145
  consensusScore: geminiConsensusScore
144
146
  });
145
147
  // Check if we've reached consensus
146
- if (geminiConsensusScore >= 8 || claudeConsensusScore >= 8) {
148
+ if (geminiConsensusScore >= CONSENSUS_THRESHOLD || claudeConsensusScore >= CONSENSUS_THRESHOLD) {
147
149
  logger.info(`Consensus reached in round ${currentRound} with score ${geminiConsensusScore}`);
148
150
  consensusReached = true;
149
151
  }
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@rlabs-inc/gemini-mcp",
3
- "version": "0.5.0",
3
+ "version": "0.5.1",
4
4
  "description": "MCP server for Gemini 3 integration with Claude Code - full frontier AI capabilities",
5
5
  "main": "dist/index.js",
6
6
  "type": "module",