Note-taking for networked thought
## Roam Research MCP Server: Networked Thought for AI-Powered Knowledge Discovery The **Roam Research MCP Server** integrates the revolutionary graph-based note-taking system directly into Google Antigravity, enabling AI assistants to navigate, query, and build upon interconnected knowledge networks. By connecting Roam's bidirectional linking and block-level references to your development workflow, this server transforms how you capture, organize, and retrieve information for complex projects. ### Why Roam Research MCP? - **Graph-Native Queries**: Navigate knowledge through relationships rather than hierarchical folders, discovering unexpected connections between concepts - **Block-Level Precision**: Reference and manipulate individual blocks of content with unique identifiers for granular knowledge management - **Daily Notes Integration**: Automatically capture development insights, decisions, and todos in date-contextualized pages - **Bidirectional Links**: Surface all references to any concept, building a web of interconnected project knowledge - **Roam API Access**: Full programmatic control over pages, blocks, queries, and graph traversal operations ### Key Features #### 1. Knowledge Graph Queries ```python from anthropic import Anthropic client = Anthropic() response = client.messages.create( model="claude-sonnet-4-20250514", max_tokens=1024, messages=[{ "role": "user", "content": "Find all blocks linked to [[API Design]] that also reference [[Security]]" }], tools=[{ "name": "roam_query", "description": "Query Roam graph for linked content" }] ) ``` #### 2. Block Management ```python # Create and link blocks programmatically response = client.messages.create( model="claude-sonnet-4-20250514", max_tokens=1024, messages=[{ "role": "user", "content": "Add a new block under [[Project Decisions]] documenting our choice to use GraphQL" }], tools=[{"name": "roam_create_block", "description": "Create blocks with references"}] ) ``` #### 3. Daily Notes Capture ```python # Append to today's daily notes response = client.messages.create( model="claude-sonnet-4-20250514", max_tokens=1024, messages=[{ "role": "user", "content": "Add today's standup notes: completed auth refactor, starting payment integration" }], tools=[{"name": "roam_daily_note", "description": "Append to daily notes page"}] ) ``` #### 4. Graph Visualization ```python # Explore connections between concepts response = client.messages.create( model="claude-sonnet-4-20250514", max_tokens=1024, messages=[{ "role": "user", "content": "Show me how [[Database Schema]] connects to [[User Stories]] through linked pages" }], tools=[{"name": "roam_graph_explore", "description": "Traverse and visualize connections"}] ) ``` ### Configuration ```json { "mcpServers": { "roam-research": { "command": "npx", "args": ["-y", "@anthropic/mcp-roam"], "env": { "ROAM_API_TOKEN": "your-roam-api-token", "ROAM_GRAPH_NAME": "your-graph-name" } } } } ``` ### Use Cases **Personal Knowledge Management**: Build and query a personal knowledge base that connects programming concepts, project learnings, and career development insights. **Project Documentation**: Create interconnected documentation where design decisions link to requirements, code references, and meeting notes. **Research Synthesis**: Aggregate information from multiple sources, automatically creating links between related concepts and surfacing patterns. **Learning Journaling**: Track learning progress with daily notes that reference concepts, questions, and breakthroughs in understanding. The Roam Research MCP Server transforms your graph database into an intelligent knowledge partner, helping you think in networks rather than files.
{
"mcpServers": {
"roam-research": {
"mcpServers": {
"roam-research": {
"env": {
"ROAM_API_TOKEN": "your-api-token",
"ROAM_GRAPH_NAME": "your-graph"
},
"args": [
"-y",
"roam-mcp-server"
],
"command": "npx"
}
}
}
}
}