agentty Alternatives

lazyagent is a monitoring tool for coding agent sessions. It is built to show what running agents are doing and to track their activity, token use, and cost in real time across Claude Code (CLI and Desktop), Cursor, Codex, Grok CLI, Kilo, Kimi Code CLI, Amp, pi, and OpenCode. The product centers on session monitoring in one place. It reports live activity states such as thinking, writing, reading, searching, and compacting, and it includes automatic token counting, USD cost estimation per session, and activity sparklines. It also provides maintenance functions from the CLI: pruning old or orphaned chats, compacting bulky session files in place, searching transcripts, and checking rate-limit windows. Transcript search spans Claude, Codex, pi, Amp, Grok, and Kimi transcripts, with highlighted snippets and an incremental local index; Cursor, OpenCode, and Kilo are excluded from that search because they keep history in third-party SQLite databases. Rate-limit summaries are available in the terminal interface and the GUI, with a quick-scan table and a detailed view showing per-window bars, reset times, and a pace indicator. lazyagent is delivered as three interfaces: a terminal UI, a macOS menu bar app, and an HTTP API. The TUI is built with bubbletea and supports Vim-style navigation, fuzzy search, activity filters, time window controls, real-time updates via FSEvents, and CJK/Unicode text rendering. The menu bar app runs silently in the background with no Dock icon, uses a Catppuccin theme, and can toggle a session panel, open the limits page, or detach from the menu bar. The HTTP API provides REST endpoints and Server-Sent Events for real-time streaming, and it is protected by passphrase-derived Bearer tokens using PBKDF2-SHA256. The API exposes session lists, session detail, stats, configuration, and an events stream, and it can be used from a browser, mobile app, or curl. Outbound webhooks can push signed HTTP POSTs to Slack, dashboards, or CI endpoints when a session changes state, with optional HMAC-SHA256 signing, event and agent filters, and automatic retries on transient failures. It is distributed through Homebrew, Go install, or source build. The project is released under the MIT license and is described as a single Go binary with no runtime dependencies and minimal CPU use. It is credited to Illegal Studio and Vincenzo Petrucci.

agent-intern is an open-source MCP server that bridges the Antigravity and OpenAI Codex CLIs, enabling Claude Code to drive them as sub-agents. It is designed for developers building complex AI agent workflows and orchestration systems. The project is MIT-licensed and supports integration with multiple LLM providers.

agentcode-cli is an open-source command-line tool that brings agentic coding capabilities inspired by Claude Code. It supports multiple AI models and is designed for developers who want to automate or enhance coding workflows using AI agents.

AgentOS is an open-source framework written in TypeScript designed for building production-ready autonomous AI agents. Developed by Frame.dev, it targets developers seeking to create adaptive agents capable of remembering, adapting, and generating new tools during runtime. The framework emphasizes emergent intelligence and supports multi-agent orchestration, enabling multiple agents to collaborate and dynamically spawn specialists when required by the task. A distinguishing feature of AgentOS is its cognitive memory system, which is grounded in neuroscience concepts such as Ebbinghaus decay and reconsolidation. This allows agents to recall and utilize information over time, contributing to their adaptability. The framework also includes multimodal retrieval-augmented generation (RAG), AI guardrails that are GDPR-ready, and voice pipeline capabilities. AgentOS supports integration with 11 large language model (LLM) providers, offering flexibility in building agents that can leverage different AI backends. Agents built with AgentOS can write their own TypeScript functions mid-task, which are then reviewed and approved by a judge agent, and new specialist agents can be spawned at runtime to expand the system’s capabilities. For development and management, AgentOS offers a desktop application called AgentOS Workbench. This tool provides a visual agent builder with drag-and-drop workflows, real-time testing and debugging tools, and options for one-click deployment to either cloud or local environments. The framework is self-hosted, allowing deployment on the user’s own infrastructure, and is distributed under the Apache 2.0 license for the core, with MIT licensing for agents, extensions, and guardrails. AgentOS has demonstrated its effectiveness with benchmark results such as 85.6% on LongMemEval-S, and it is used as the foundation for Paracosm, an AI agent swarm simulation engine that enables scenario-based simulations with AI commanders exhibiting distinct personalities. Extensive documentation, code examples, and community support are available through official channels, including Discord and GitHub.

Agent Hub is an open-source native macOS application designed to unify multiple AI agents, including Claude Code, Codex, Hermes, and OpenClaw, within a single interface. It allows users to interact with these agents either locally on their Apple Silicon Mac or remotely over SSH on any accessible server or VPS. The platform is aimed at users who prefer to keep their existing code editors, offering a separate chat-based client for agent interaction rather than integrating directly into an IDE. The tool brings together the official agent command-line interfaces, providing a chat window where users can send commands and receive responses from any supported agent. Agent Hub supports both local and remote execution modes for each agent, with the ability to connect to any OpenAI-compatible endpoint, including local instances or hosted APIs. Users can switch between different agents and remote hosts, manage granular permissions per agent, and approve or deny specific tool calls. The interface features slash commands for quick access to agent capabilities, fuzzy command search, and supports file and image uploads directly into the chat window. Model selection is available per thread, enabling users to choose between various versions and sizes of supported models. Agent Hub is delivered as a downloadable macOS application for Apple Silicon devices running macOS 11 or later. It is also available as source code on GitHub, allowing users to clone, build, and run the app themselves. API keys are stored securely in the OS keychain, either locally or on the remote host, and the tool does not transmit telemetry or analytics data. The application is distributed under the MIT license, making it free to use and modify, with users responsible for any costs associated with their chosen AI model providers. This platform is particularly suited for developers, terminal users, and those managing multiple AI agents across local and remote environments, providing a unified and flexible interface for agent-driven workflows.

Terminal-agent runtime for LLM-driven interactive TTY sessions

agentproto CLI is a command-line tool designed to install, run, and orchestrate agent CLIs and adapters. It serves as the reference host for AgentProto adapters, allowing users to run individual turns, maintain sessions, serve as a daemon, and manage swarms of agents. The tool is distributed as a single binary and can be installed globally using npm. Key features include the ability to run agents locally, either for a single turn or as persistent sessions, and to install various adapters such as claude-code and hermes. Sessions are managed in a live dashboard panel, which provides a session sidebar and multiple views per session, including terminal, chat, JSON, and event views. The panel supports shareable deep-links over a tunnel, enabling remote access and collaboration. agentproto CLI can detect and communicate with daemons on local or remote ports within a specified range. The tool supports exposing a machine as a daemon over an outbound WebSocket tunnel after logging in via device flow. It offers swarm orchestration by composing a substrate, dispatcher, and participants as defined in a manifest, with kernel routing and plugin extensibility. The daemon functions as an MCP server, providing HTTP transport for Claude Code and includes a bundled stdio bridge. agentproto CLI is suitable for users who need to manage and orchestrate multiple agent CLIs and adapters, providing both local and remote management capabilities through its command-line interface and live dashboard panel.

agent-container is an open-source tool that enables developers to create always-on, containerized development environments for AI coding agents. It leverages SSH and tmux for remote access and persistent sessions, supporting Docker and Podman. Ideal for developers needing robust, automated agent workspaces.

Agent of Empires is a terminal session manager for AI coding agents. It is built for managing multiple Claude Code, OpenCode, and similar agent sessions, with a TUI for monitoring and controlling them from the terminal. The product also uses tmux for persistence, so sessions keep running if aoe is closed. Its features include a visual TUI dashboard for viewing session status and navigating with keyboard shortcuts. It supports Git worktrees for running parallel agents on different branches of the same repository, with automatic worktree creation and cleanup. Docker sandboxing is available for running agents in isolated containers, and the listed controls include configurable volume mounts, persistent authentication, and automatic lifecycle management. The tool also provides status detection for agents that are working, waiting for input, or idle, without requiring session switching. Agent of Empires also offers a web dashboard that can drive the same sessions from any browser. The page includes a real terminal in the browser, PWA install support on desktop and mobile, and a structured view that renders agent state through the Agent Client Protocol, including plan panels, tool-call cards, and swipe-to-approve interactions. Remote phone access is supported by exposing the dashboard over HTTPS with QR pairing, and the named connectivity options are Tailscale Funnel and Cloudflare Tunnel. It also supports multi-repo workspaces through a project registry and multi-select pickers. The site lists Homebrew and an install script for installation. It identifies the project as built with Rust and powered by tmux, and it is distributed under the MIT License. The page also names tutorials, walkthroughs, setup guides, and help videos as learning resources.

agent-cli-Mrzhou300 is an open-source, lightweight command-line personal assistant that integrates design concepts from multiple open-source agent projects. It leverages AI models, including Anthropic's Claude, to automate tasks and assist users directly from the terminal. Ideal for developers and technical users.

agent-babysitter is an open-source CLI tool that enables developers to monitor Claude Code agents and receive real-time updates on their phone via Telegram. It is designed for developers who want to keep track of agent activity while away from their desk.

oragent is an open-source CLI tool that acts as a supervisor for AI coding agents, enabling users to manage multiple Claude Code, Codex, and Shell sessions in parallel. It provides a state-aware cockpit for efficient agent orchestration.