Gemini

Connect Crow to Google's Gemini — both the Gemini CLI and Gemini in Google AI Studio/Enterprise.

Gemini CLI — Local (stdio)

Best for development — runs Crow servers directly on your machine. No gateway required.

Setup Steps

1. Clone and set up Crow locally:

   git clone https://github.com/kh0pper/crow.git
   cd crow
   npm run setup

2. Edit ~/.gemini/settings.json:

   {
     "mcpServers": {
       "crow-memory": {
         "command": "node",
         "args": ["/path/to/crow/servers/memory/index.js"],
         "env": {
           "CROW_DB_PATH": "/path/to/crow/data/crow.db"
         }
       },
       "crow-projects": {
         "command": "node",
         "args": ["/path/to/crow/servers/research/index.js"],
         "env": {
           "CROW_DB_PATH": "/path/to/crow/data/crow.db"
         }
       }
     }
   }

3. Restart Gemini CLI — it will detect the MCP servers automatically.

TIP

Run npm run mcp-config in the Crow directory to generate a complete MCP config, then copy the relevant entries to your Gemini settings.

Gemini CLI — Remote (HTTP)

Connect to a deployed Crow gateway for the full platform including external integrations.

Prerequisites

• Crow gateway deployed and healthy (Cloud Deploy Guide)

Setup Steps

1. Edit ~/.gemini/settings.json:

   {
     "mcpServers": {
       "crow-memory": {
         "url": "https://your-crow-server/memory/mcp"
       },
       "crow-projects": {
         "url": "https://your-crow-server/projects/mcp"
       },
       "crow-tools": {
         "url": "https://your-crow-server/tools/mcp"
       }
     }
   }

2. Start Gemini CLI — it will auto-discover the OAuth metadata and prompt for authorization.

Google AI Studio

Google AI Studio supports MCP servers for tool use in chat. Configuration is available through the interface when creating a new chat or agent.

Setup Steps

1. Open Google AI Studio
2. Create a new chat or agent
3. In the tools section, add an MCP server
4. Enter the Streamable HTTP URL:

   https://your-crow-server/memory/mcp

5. Complete the OAuth authorization flow

Self-Hosted / Local Setup

If you're running the Crow gateway on your own machine, you can expose it to Gemini using Tailscale Funnel. Once Funnel is enabled on the machine running the gateway, your MCP endpoint URL will be:

https://<hostname>.<tailnet>.ts.net/memory/mcp

Replace <hostname> and <tailnet> with your Tailscale machine name and tailnet domain. Use the same URL pattern for other servers (/projects/mcp, /router/mcp, etc.). The setup steps are identical to the remote instructions above — just substitute your Funnel URL for your-crow-server in the Gemini settings.

See the Tailscale Setup guide for full configuration details.

Transport

• Type: Streamable HTTP
• Protocol: 2025-03-26
• Auth: OAuth 2.1 (automatic discovery)

Cross-Platform Context

Crow automatically delivers behavioral context when Gemini connects — memory protocols, session management, and transparency rules are active from the first message. No manual loading required.

For detailed guidance, Gemini can request MCP prompts like session-start, crow-guide (with platform: "gemini"), or feature-specific guides (research-guide, blog-guide, sharing-guide).

You can also manually load the full context:

"Use the crow_get_context tool with platform set to gemini"

Memories and projects stored from any platform are shared. See the Cross-Platform Guide.

Verification

Ask Gemini:

"Store a memory that Gemini is connected to Crow."

Then verify:

"Search memories for 'Gemini'."