HubLensLLMrowboatlabs/rowboat
// archived 2026-04-12
rowboatlabs

rowboat

AI#Knowledge Graph#Local-first#LLM#Markdown#Productivity
View on GitHub
109

// summary

Rowboat is an open-source AI coworker that integrates with your email and meeting notes to build a persistent, local knowledge graph. It utilizes this context to assist with tasks like drafting documents, preparing for meetings, and tracking projects while keeping all data in an editable Markdown format. By maintaining long-lived memory on your machine, it allows for compounding context that improves over time without relying on external cloud storage.

// technical analysis

Rowboat is a local-first AI coworker designed to transform fragmented work data into a persistent, structured knowledge graph. By storing information as an Obsidian-compatible vault of plain Markdown files, it ensures that user context is transparent, editable, and portable, avoiding the limitations of opaque, on-demand retrieval systems. This architecture prioritizes data sovereignty and compounding memory, allowing the AI to act on historical context while maintaining privacy by keeping all data on the user's machine.

// key highlights

01
Maintains a long-lived knowledge graph using plain Markdown files that are fully inspectable and editable by the user.
02
Integrates with Gmail, Google Calendar, and meeting note services to automatically build context from existing workflows.
03
Supports a 'bring your own model' approach, allowing users to switch between local models via Ollama/LM Studio or hosted API providers.
04
Features 'Live Notes' that automatically update based on tracked topics, people, or projects across various data sources.
05
Extends functionality through the Model Context Protocol (MCP), enabling connections to external tools like Slack, Jira, and web search.
06
Enables the creation of tangible artifacts such as PDF decks and meeting briefs by leveraging the accumulated knowledge graph.

// use cases

01
Generate meeting briefs and PDF slide decks based on historical decisions and context.
02
Maintain live notes that automatically update to track specific people, projects, or market topics.
03
Extend functionality through the Model Context Protocol (MCP) to connect with external tools like Slack, Jira, and web search.

// getting started

To begin using Rowboat, download the latest installer for your operating system from the official website or GitHub releases page. After installation, configure your desired integrations by following the setup guides for Google services and adding necessary API keys for optional features like voice input or web search in the ~/.rowboat/config/ directory. Once set up, the application will begin building your local knowledge vault from your connected accounts.