vlinx-io/VelaTerm

What it solves

VelaTerm is an Agentic Development Environment (ADE) that centralizes the management of multiple AI coding agents, terminal sessions, and development workflows into a single, unified interface. It solves the fragmentation of running various agents (like Claude Code, Copilot, or Cursor) across different windows by providing a structured, persistent workspace that works across desktop, browser, and mobile.

How it works

VelaTerm acts as a management layer for AI agent sessions. Each agent runs in its own managed pseudo-terminal session, which can be organized into a hierarchical tree of projects and groups. The platform allows agents to communicate with each other, delegate sub-tasks to child sessions, and access a shared knowledge base. It provides a split-pane terminal interface, real-time status monitoring, and remote access capabilities via SSH or HTTPS.

Who it’s for

This tool is designed for developers who use multiple AI coding agents simultaneously and need a way to orchestrate complex, multi-agent workflows, manage long-running background tasks, and maintain access to their development environment from any device.

Highlights

• Multi-Agent Orchestration: Supports "Plan/Execute" modes where a planner agent delegates tasks to multiple executor agents in parallel.
• Cross-Session Communication: Agents can search, reference, and message other sessions using built-in commands like vsearch, vrefer, and vtell.
• Unified Interface: Combines terminal management, conversation views, file editing, and knowledge base lookups into one native application.
• Anywhere Access: Provides consistent access to your session tree via desktop apps (macOS, Windows, Linux), web browsers, and mobile devices.
• Extensible Knowledge: Includes a session-aware knowledge base (vkb) that allows agents to query notes and code graphs to improve task context.