BolideAI MCP
A comprehensive ModelContextProtocol server that provides AI-powered tools for marketing automation, content generation, research, and project management, integrating with various AI services to streamline workflows for developers and marketers.
README
BolideAI MCP
BolideAI MCP is a comprehensive ModelContextProtocol (MCP) server that provides tools for marketing automation, content generation, research, and project management. It integrates with various AI services to streamline workflows for developers and marketers.
Features
- 🚀 Project Scaffolding - Create marketing project directories and structures
- 📱 Marketing Automation - Capture screenshots and videos using companion app
- 🎯 Artifact Management - Organize marketing materials with structured directories
- 🤖 AI-Powered Content Generation - Generate social media posts using Gemini AI
- 🔍 Research Tools - Comprehensive research using Perplexity AI and OpenAI
- 📊 Asset Management - Organize and manage marketing assets
- 🛠️ Diagnostic Tools - System validation and troubleshooting
Getting Started
Quick Start (NPM Package)
For quick testing without local builds:
Installation
npm install -g @bolide-ai/mcp
Or use with npx (recommended):
npx @bolide-ai/mcp@latest
Configuration
Configure your MCP client with the following:
{
"mcpServers": {
"BolideAI": {
"command": "npx",
"args": ["-y", "@bolide-ai/mcp@latest"],
"env": {
"BOLIDEAIMCP_API_KEY": "your-api-key",
}
}
}
}
Development Setup (Local Build)
For development or local builds, follow these detailed steps:
1. Install Prerequisites
Install Node.js 22+
# Download and install from https://nodejs.org/en/download
Clone the repositories
git clone https://github.com/Bolide-AI/mcp
2. Build the MCP Server
# In the directory of the cloned repository
npm install && npm run build
4. Configure MCP in Cursor
- In Cursor menu select Cursor → Settings → Cursor Settings
- In the opened window select Tools & Integrations
- Click New MCP Server
- Insert the MCP server configuration, replacing:
<PATH TO MCP DIRECTORY>with the path to the MCP directory<BOLIDEAI_API_TOKEN>with your BolideAI key (can be generated at here)
{
"mcpServers": {
"BolideAI-dev": {
"type": "stdio",
"command": "node",
"args": [
"--inspect=9999",
"<PATH TO MCP DIRECTORY>/build/index.js"
],
"env": {
"BOLIDEAI_MCP_DEBUG": "true",
"BOLIDEAI_API_TOKEN": "<your-api-key-here>"
}
}
}
}
- Make sure the workspace is open in Cursor
- Launch the application in the simulator
Start Using Tools
// Create a project
scaffold_marketing_project()
// Perform research
use_openai_deep_research({
query: "AI trends in marketing automation 2024"
})
// Analyze video content
analyze_videos({
artifactName: "my-project-20240101",
videoNames: ["demo.mov"],
force: false
})
Environment Variables
BOLIDEAI_API_TOKEN- RequiredBOLIDEAI_API_URL- Optional, defaults to https://bolide.ai/api
Tool Configuration
BOLIDEAIMCP_DEBUG=true- Enable diagnostic tools and detailed logging- Tool Groups - Enable specific tool categories (see Tool Options)
- Individual Tools - Enable specific tools only (see Tool Options)
Available Tools
BolideAI MCP provides 15 tools across 7 categories:
🚀 Project Scaffolding
scaffold_marketing_project- Create marketing project directory structure
📱 Utility Tools
check_companion_app_status- Check companion app running statuslaunch_companion_app- Launch Companion App for marketing capturestop_companion_app- Stop running companion app instances
🎯 Artifact Management
create_artifact_directory- Create organized artifact directoriescreate_post_artifact- Store post artifacts with metadata
📊 Asset Management
create_post_asset- Create marketing asset filescreate_research_asset- Create research asset files
🤖 Content Generation
analyze_videos- Analyze video content using AIgenerate_gif- Convert video segments to GIFsenhance_audio- Extract and enhance audio from videos using ElevenLabsfetch_reddit_posts- Fetch Reddit posts from specified subreddits
🔍 Research Tools
use_perplexity- Research using Perplexity AIuse_openai_deep_research- Deep research using OpenAI o4-mini-deep-research
🛠️ Diagnostic Tools
diagnostic- System environment validation (debug mode only)
Research Tools
Perplexity AI Research
Perform quick research and information gathering:
use_perplexity({
query: "Latest trends in AI marketing automation",
search_mode: "web" // or "academic"
})
OpenAI Deep Research
Conduct comprehensive research with query enrichment:
use_openai_deep_research({
query: "Economic impact of renewable energy adoption"
})
The deep research tool:
- Enriches your query using GPT-4.1 with detailed research instructions
- Researches using o4-mini-deep-research with web search and code interpreter
- Returns both enriched instructions and comprehensive findings
Common Workflows
Complete Marketing Content Creation
-
Set up project structure:
scaffold_marketing_project() -
Create artifact directory:
create_artifact_directory({ artifactName: "feature-demo" }) -
Check app status and capture content:
check_companion_app_status() launch_companion_app({ artifactsDirectory: "/path/to/artifacts/feature-demo-20240101" }) -
Enhance audio quality:
enhance_audio({ artifactName: "feature-demo-20240101", videoNames: ["demo.mov"] }) -
Store results:
create_post_artifact({ artifactName: "feature-demo-20240101", fileName: "final-post.md", fileContent: "Generated content..." }) -
Store research:
create_research_asset({ name: "market-research.md", content: "Detailed research findings..." })
Research Workflow
-
Quick research:
use_perplexity({ query: "Your research question", search_mode: "web" }) -
Deep research:
use_openai_deep_research({ query: "Complex research topic requiring comprehensive analysis" }) -
Store research findings:
create_research_asset({ name: "market-analysis.md", content: "Research findings and analysis..." })
Reddit Data Collection
-
Fetch Reddit posts:
fetch_reddit_posts({ subreddit: "programming", limit: 25, sort: "top", timePeriod: "day" }) -
Store Reddit data:
create_research_asset({ name: "reddit-programming-trends.md", content: "Analysis of top programming posts..." })
Configuration Options
Selective Tool Registration
Enable only the tools you need to optimize performance:
{
"env": {
"BOLIDEAIMCP_GROUP_RESEARCH": "true",
"BOLIDEAIMCP_GROUP_CONTENT_GENERATORS": "true",
"GOOGLE_API_KEY": "your-api-key",
"OPENAI_API_KEY": "your-api-key"
}
}
Available Tool Groups
BOLIDEAIMCP_GROUP_LAUNCH- Launch and utility toolsBOLIDEAIMCP_GROUP_SCAFFOLDING- Project scaffolding toolsBOLIDEAIMCP_GROUP_ARTIFACTS- Artifact management toolsBOLIDEAIMCP_GROUP_ASSET_GENERATORS- Asset management toolsBOLIDEAIMCP_GROUP_CONTENT_GENERATORS- Content generation toolsBOLIDEAIMCP_GROUP_RESEARCH- Research and information gathering toolsBOLIDEAIMCP_GROUP_DIAGNOSTICS- Diagnostic tools
Documentation
- Tool Reference - Comprehensive tool documentation
- Tool Options - Configuration and selective tool registration
- Contributing - Development and contribution guidelines
System Requirements
Dependencies
- Node.js 18+ (automatically handled by npx)
- ffmpeg (required for GIF generation tools)
- Companion App
Troubleshooting
Enable Debug Mode
{
"env": {
"BOLIDEAIMCP_DEBUG": "true"
}
}
Run Diagnostics
diagnostic() // Available in debug mode
Common Issues
- Missing API Keys: Ensure all required environment variables are set
- ffmpeg Not Found: Install ffmpeg using
brew install ffmpeg - Permission Issues: Check file system permissions for artifact directories
Contributing
We welcome contributions! Please see our Contributing Guide for details.
License
MIT © Data Route LLC
Support
- 🐛 Report Issues
- 💬 Support
BolideAI MCP - Streamline your marketing automation and research workflows with AI-powered tools.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。