SuperAPI AI Directory
Partner with us中文AI newsFavorites

AI agents

2,963 results · Page 26 of 83

VibeBox

AI agents

VibeBox is an ultra-fast open-source AI sandbox that provides developers with a secure, isolated virtualized environment. Its core capabilities include: hot start time under 6 seconds, using cached base images to achieve near-instant re-entry; agents can run tasks and patch files within the virtual machine, with complete command freedom; isolation by design based on Apple Virtualization, completely sealing off the host machine; predictably setting CPU, memory, disk size, and mounts via vibebox.toml; sessions are auditable, and project state is saved in .vibebox, including logs and keys; entering a ready sandbox with just one command. This tool is aimed at developers who need to run AI agents quickly and securely, emphasizing a developer-first design philosophy.

Read more

vicoa

AI agents

Vicoa is an open-source AI orchestrator (agentic IDE) for running a team of coding agents from a desktop, mobile device, or VPS. It supports more than 40 coding agents, including Claude Code, Codex, Cursor, Gemini, GitHub Copilot, Kimi, and others, and can run them in parallel in the same workspace. Each agent has its own independent git worktree and branch, avoiding interference with one another. Users can view the status of all sessions through a unified command center and choose where to run them on Mac, Windows, Linux, a VPS, or a remote server. Vicoa uses a bring-your-own-key (BYO-key) model, using the user's existing subscription or API key. The mobile side provides native iOS and Android apps, supporting notifications, @file mentions, slash commands, voice dictation, inline diffs, file browsing, a terminal, and live mobile preview. It also provides task management, automation (cron scheduling), skill management, and CLI support. Vicoa is fully open source, can be self-hosted, uses Docker to deploy the entire stack, and is licensed under AGPLv3.

Read more

Voyager

AI agents

Voyager is the first LLM-powered embodied lifelong learning agent in Minecraft, capable of continuously exploring the world, acquiring diverse skills, and making new discoveries without human intervention. It consists of three core components: an automatic curriculum that maximizes exploration, an executable code skill library that stores and retrieves complex behaviors, and an iterative prompting mechanism that combines environmental feedback, execution errors, and self-verification. Voyager interacts with GPT-4 through black-box queries without fine-tuning model parameters. Its skills are temporally extended, interpretable, and compositional, enabling rapid compounding of the agent's abilities and mitigating catastrophic forgetting. Experiments show that Voyager obtains 3.3 times more unique items, travels 2.3 times farther, and unlocks key tech tree milestones up to 15.3 times faster than previous state-of-the-art methods. Voyager can use its learned skill library to solve new tasks from scratch in new worlds. This repository provides the code under the MIT license and requires Python >= 3.9 and Node.js >= 16.13.0, and depends on the Minecraft game, Fabric mods, and an OpenAI API key.

Read more

Vrizen

AI agents

Vrizen is an engineering team that provides custom web applications, workflow automation, and AI agents, using a fixed-price model, with code ownership belonging to the client. The team is located in Chennai, India, and serves clients globally. Its core capabilities include: building AI assistants based on proprietary documents for customer support, research, and internal knowledge workflows; building automated workflows with error handling, visibility, and clear human handoff; and developing web applications based on mainstream, maintainable infrastructure. Its product Reply is an AI customer support agent that supports WhatsApp, Instagram, Telegram, and websites, with answers based on the client's own approved documents. Vrizen emphasizes from strategy to delivery: first mapping out people, tools, data, and decisions, then proposing technical solutions, and clarifying processes, interfaces, integrations, risks, and ownership before quoting.

Read more