awesome-repositories.com
Blog
MCP
awesome-repositories.com

Discover the best open-source repositories with AI-powered search.

ExploreCurated searchesOpen-source alternativesSelf-hosted softwareBlogSitemap
ProjectMCP serverAboutHow we rankPress
LegalPrivacyTerms
© 2026 Bringes Technology SRL·VAT RO45896025·hello@awesome-repositories.com
vastsa avatar

vastsa/PI-Desktop

0
View on GitHub↗
pi-docs.aiuo.net
↗

PI Desktop

PI-Desktop is a desktop application that runs an AI coding agent on your own machine against a local repository. Projects, sessions, settings, and logs are stored locally, credentials are kept in the system keychain, model requests go straight to your chosen provider, and no telemetry is collected. It is a persistent workspace, independent of any particular editor or terminal, that remembers its state between launches. It is available for macOS, Windows, and Linux.

The agent can work in agent, plan, or goal modes: delegate a coding task and review progress, study a project and approve an implementation plan before anything changes, or set an objective with acceptance criteria and let the agent keep going until it is met. Larger jobs can be split across background subagents or full worker sessions, each with independent context, coordinated from one parent session with inspectable transcripts. Privileged actions such as file edits, command execution, and tool calls pass through a permission layer, so you decide how much autonomy each session gets.

The app is model-agnostic. You can connect cloud models, local runtimes, custom gateways, and compatible APIs, and switch between them mid-session with per-model settings for context size, output limit, reasoning, and credentials. It also hosts Model Context Protocol servers, local or remote, whose tools become available to the agent during a session. Plugins extend the workspace with panels, agent tools, skills, themes, floating widgets, background services, and MCP servers. Built-in templates scaffold new plugins, and they can be packaged as portable files or shared through a marketplace.

Day-to-day work is organized into projects and sessions that can be pinned, archived, branched, searched, queued, and resumed, with checkpoints for recovering interrupted work. A side panel shows file diffs, command output, and agent activity, with @ mentions for project files and slash commands for actions. Sessions from other coding agents can be imported to continue earlier work. Desktop packages come for common architectures, with signed and notarized macOS releases.

Features

  • AI Coding Agent Platforms - A desktop application that runs an AI coding agent locally, keeping projects, sessions, and credentials on the user's machine.
  • AI Agents - Runs an AI coding agent on your own machine against a local repository, with agent, plan, and goal modes.
  • Agent Task Assignment - Lets you hand a coding task to an agent that reads code, edits files, runs commands, and tests while you review progress.
  • Agentic Workspaces - Provides a persistent desktop workspace combining projects, sessions, reviews, previews, and agents, independent of any editor or terminal.

AI search

Explore more awesome repositories

Describe what you need in plain English — the AI ranks thousands of curated open-source projects by relevance.

Start searching with AI
7,770 stars·762 forks·TypeScript·LGPL-3.0·3 views
  • Local Data Privacy Tools - Stores projects, sessions, settings, and logs locally, keeps credentials in the system keychain, sends requests directly to providers, and collects no telemetry.
  • AI Agent Permissions - Routes file edits, command execution, and tool calls through a permission layer so users set how much autonomy each session gets.
  • Sub-Agent Task Delegation - Spawns background subagents with their own context for exploration, implementation, testing, research, or review, reporting back to the parent agent.
  • Agent Goal Definitions - Supports a goal mode where you set an objective with acceptance criteria and the agent keeps working until it is met.
  • Plugin-Extensible Agent Interfaces - Lets users extend the agent workspace with installable plugins that add panels, agent tools, skills, themes, floating widgets, and background services.
  • Coding Agent Orchestrators - Coordinates background subagents and full worker sessions side by side from one parent session, each with independent context and inspectable transcripts.
  • Multi-Agent Orchestration Platforms - Coordinates background subagents and full worker sessions from one parent session, each keeping independent context and an inspectable transcript.
  • MCP Server Connections - Connects local or remote Model Context Protocol servers so their tools become available to the agent during a session.
  • Model Context Protocol Integrations - Hosts local or remote Model Context Protocol servers whose tools become available to the agent during a coding session.
  • Model Provider Configurations - Lets users configure model providers, credentials, and per-model settings so the coding agent runs on their own models.
  • Runtime Provider Switching - Switches between cloud models, local runtimes, and custom gateways mid-session with per-model settings, avoiding vendor lock-in.
  • Model Provider Management - Connects cloud models, local runtimes, custom gateways, and compatible APIs and manages switching between them without changing workflow.
  • Provider-Agnostic Model Interfaces - Connects to cloud models, local runtimes, and custom gateways through one client, so users can switch providers without vendor lock-in.
  • Runtime Model Switching - Allows switching the active model or provider mid-session, with per-model settings for context size, output limit, reasoning, and credentials.
  • Coordinator-Worker Orchestration - Delegates longer-lived work to full worker sessions with independent context, managed from one parent session with inspectable transcripts.
  • Session Branching - Organizes work into projects and sessions that can be pinned, archived, branched, searched, queued, and resumed, with checkpoints for recovery.
  • Agentic Plugin Marketplaces - Installs add-ons from a plugin marketplace to widen the agent's abilities and change how the desktop app works.
  • Plugin Extenders - Supports installing plugins or building your own to add new capabilities to the desktop workspace.
  • Parallel Subagent Orchestrators - Splits large tasks across parallel subagents or full worker sessions and gathers their results into one parent session.
  • Implementation Planning - Offers a plan mode that studies the project and presents an implementation plan for approval before any edit runs.
  • Plan Approval Workflows - Offers a plan mode that studies the project and presents an implementation plan for approval before any change is made.

    • Star history

    Star history chart for vastsa/pi-desktopStar history chart for vastsa/pi-desktop

    How this analysis was created: This summary and feature list are AI-generated from collected project material and can contain mistakes. Stars, license and language are imported from GitHub. Inclusion does not mean that we have tested or audited this project. Check the source documentation for any feature you depend on. Learn more on our About page.

    Frequently asked questions

    What does vastsa/pi-desktop do?

    PI-Desktop is a desktop application that runs an AI coding agent on your own machine against a local repository. Projects, sessions, settings, and logs are stored locally, credentials are kept in the system keychain, model requests go straight to your chosen provider, and no telemetry is collected. It is a persistent workspace, independent of any particular editor or terminal, that remembers its state between launches. It is available for macOS, Windows, and Linux.

    What are the main features of vastsa/pi-desktop?

    The main features of vastsa/pi-desktop are: AI Coding Agent Platforms, AI Agents, Agent Task Assignment, Agentic Workspaces, Local Data Privacy Tools, AI Agent Permissions, Sub-Agent Task Delegation, Agent Goal Definitions.

    Which projects share features with vastsa/pi-desktop?

    Projects with overlapping indexed features include: claude-code-best/claude-code — Claude Code is a command-line interface and multi-agent orchestration framework designed for autonomous software… fathah/hermes-desktop — Hermes Desktop is a cross-platform desktop application that acts as a graphical companion for installing, configuring,… liuup/claude-code-analysis — This repository is a static analysis of the Claude Code terminal coding assistant, based on a reading of its… lintsinghua/claude-code-book — This repository is a long-form technical book about how a production AI coding agent runtime, often called an agent… builderz-labs/mission-control — Mission Control is a self-hosted web dashboard that acts as a control plane for autonomous AI coding agents. It lets… docker/docker-agent — This project is a container-native runtime designed for building, orchestrating, and executing autonomous AI agents.…

    Projects sharing features with PI Desktop

    These projects share indexed features with PI Desktop. Shared tags can include platform or build tooling; verify the primary use case before treating a result as a replacement.
    • claude-code-best/claude-codeclaude-code-best avatar

      claude-code-best/claude-code

      20,272View on GitHub↗

      Claude Code is a command-line interface and multi-agent orchestration framework designed for autonomous software engineering. It enables AI agents to perform codebase modifications, debugging, and Git workflow management while coordinating multiple specialized agents to decompose and execute complex engineering tasks in parallel. The system distinguishes itself through a high degree of isolation and safety, utilizing Git worktrees to create independent working directories for concurrent agents and implementing a tiered permission system that combines user rules, project policies, and OS-level

      TypeScript
      View on GitHub↗20,272
    • fathah/hermes-desktopfathah avatar

      fathah/hermes-desktop

      14,548View on GitHub↗

      Hermes Desktop is a cross-platform desktop application that acts as a graphical companion for installing, configuring, chatting with and managing a Hermes AI agent. The agent can run locally on the user's machine or on a remote machine, reached through an API server URL and key or through SSH tunnels. A streaming chat interface shows replies, reasoning, tool progress and live token and cost counts as they arrive, supports slash commands, and lets users search and resume past sessions. Beyond chat, the app works as a management console for the agent. Users can create and switch between isolate

      TypeScriptaiai-agentautonomous-agents
      View on GitHub↗14,548
    • liuup/claude-code-analysisliuup avatar

      liuup/claude-code-analysis

      4,069View on GitHub↗

      This repository is a static analysis of the Claude Code terminal coding assistant, based on a reading of its TypeScript source. It documents how the agent is built, from its startup path and a single shared execution core to the streaming tool-call loop, sub-agents, and terminal interface layers. It is a reference and study resource about the agent rather than the agent itself. The analysis centers on a few areas. One is the shared core that routes queries, sub-agents, tools, and permission checks through the same control flow. Another is memory: layered, file-based records that users can ins

      TypeScript
      View on GitHub↗4,069
    • lintsinghua/claude-code-booklintsinghua avatar

      lintsinghua/claude-code-book

      4,359View on GitHub↗

      This repository is a long-form technical book about how a production AI coding agent runtime, often called an agent harness, is built. It dissects a real-world coding agent's architecture chapter by chapter and turns it into transferable lessons, and it also works as a step-by-step guide to designing and building your own harness around a language model. Python scripts in the repository support the book itself rather than the agent. The book's core subject is the machinery around the model: a streaming conversation loop that repeats tool calls until the model returns plain text, a tool protoc

      Pythonagentagent-architectureai-agent
      View on GitHub↗4,359
    Compare all 30 related projects→