siddsachar/row-bot

Row-Bot - Personal AI Sovereignty. A local-first AI assistant with integrated tools, a personal knowledge graph, voice, vision, shell, browser automation, scheduled tasks, health tracking, and messaging channels. Run locally via Ollama or add opt-in cloud models. Your data stays on your machine.

What it solves

Row-Bot is a local-first desktop AI assistant designed for complex, multi-step work. It addresses the limitation of simple chat interfaces by providing a system that can reason through messy context, orchestrate multiple specialized agent profiles, and directly interact with local files, repositories, and external tools.

How it works

Row-Bot operates as a desktop application that allows users to choose their preferred model path—whether local (via Ollama), hosted providers (like OpenAI or Anthropic), or custom OpenAI-compatible endpoints. It uses a LangGraph ReAct agent architecture to manage tasks. For larger projects, it employs a "Goal Mode" where a parent agent orchestrates scoped child agents for specific subtasks like research or implementation.

To maintain performance in long conversations, it uses context metering and rolling compaction to summarize older turns into reference context. It also features a personal knowledge graph for durable memory and a plugin system for extending its capabilities through MCP (Model Context Protocol) and custom tools.

Who it’s for

It is designed for developers, designers, and power users who need an AI assistant that can manage local workspaces, automate browser and computer use, and handle complex workflows with durable memory and agent orchestration.

Highlights

  • Agent Orchestration: Supports parent-led child agents, Goal Mode, and folder-scoped parallel writers for concurrent work in local directories.
  • Extensive Tooling: Includes 30+ core tools for web search, filesystem access, shell execution, and managed browser automation.
  • Developer & Designer Studios: Dedicated environments for Git workspace linking, code threads, and interactive design runtimes for mockups and landing pages.
  • Local-First Privacy: App data stays local by default, with no first-party telemetry or hosted account system.
  • Broad Model Support: Compatible with a wide range of providers including Ollama, OpenAI, Anthropic, Google AI, and various self-hosted engines like vLLM and llama.cpp.

Related

  • Project
    shiwenwen/hope-agent

    A local-first desktop AI agent that autonomously pursues long-term goals, manages a personal knowledge base, and controls the computer and browser to execute complex tasks.

  • Project
    rowboatlabs/rowboat

    A desktop AI coworker that indexes work data into a local-first knowledge graph to provide persistent memory and integrated work surfaces for AI collaboration.

  • Project
    hAcKlyc/MyAgents

    An open-source desktop workspace for personal AI agents that integrates local files, terminals, and a task management system to move beyond simple chat interfaces.

  • Project
    netease-youdao/LobsterAI

    A desktop AI agent that automates office workflows by interacting with local files, terminal commands, and browser automation, featuring remote control via IM platforms.

  • Project
    AQBot-Desktop/AQBot

    A local-first desktop AI workbench that unifies multiple LLM providers, AI agents, knowledge bases, and MCP tools into a single, privacy-focused interface.