30+ apps · one protocolConnect ContextBridge to any AI Add the ContextBridge MCP server once (it installs from this site with npx, no npm package needed). Give it your memory key from the Memory page and your agents can search and add to the same keyed memory. Notes they add are also bridge pages anyone with the link can open. Pick your app for exact setup steps: the popular ones are up top, the long tail is all there too.
Popular
Claude Desktop Anthropic Native MCP Claude Code Anthropic Native MCP Cursor Anysphere Native MCP VS Code (Copilot) Microsoft Native MCP ChatGPT OpenAI Action / API Gemini Google Copy-paste Windsurf Codeium Native MCP Cline Cline Native MCP
Everything else, connect anything
Kilo Code Roo Code Continue Zed Goose JetBrains AI Jan LibreChat Cherry Studio Witsy 5ire Warp Copilot Studio DeepSeek Grok Le Chat Perplexity Poe HuggingChat Ollama / Open WebUI LM Studio Msty
Claude Desktop Native MCP Anthropic
Speaks the Model Context Protocol directly. Add the server once and the agent can bridge links and, with your memory key, read and write your memory.
Open Settings → Developer → Edit Config (claude_desktop_config.json). Paste the ContextBridge MCP server block into the mcpServers object. Optional: set CONTEXTBRIDGE_KEY to a key from the Memory page (/memory) so the memory tools search your own memories. Save and restart the app. Ask the agent: “use contextbridge to load <paste a ChatGPT share link>”. MCP server config Copy
{
"mcpServers": {
"contextbridge": {
"command": "npx",
"args": [
"-y",
"https://contextbridge.in/downloads/contextbridge-mcp-0.2.0.tgz"
],
"env": {
"CONTEXTBRIDGE_URL": "https://contextbridge.in",
"CONTEXTBRIDGE_KEY": "cbk_..."
}
}
}
}CONTEXTBRIDGE_KEY is optional. It unlocks search_memory, add_memory and list_memories for your own memories. Create a key on the Memory page (/memory) and paste it in place of cbk_... . Without it, bridge_share_link and get_context still work. Create a memory key .
Official docs ↗