Tool profile · Productivity

open-tag

An open, self-hosted workspace built for people and AI agents to work together.

An open, self-hosted workspace built for people and AI agents to work together.

Team collaborationAI agents

01 / Overview

About

open-tag positions itself as an open workspace for people and agents. Channels, threads, and direct messages share context; agents can take tasks in their own workspace, edit files, run commands, and report back, with the control and compute planes supported by machines you control.

02 / Core features

Core features

01

Unified workspace

Channels, threads, and direct messages share context and history.

02

Mention to delegate

@ an agent to take work, edit files, run commands, and report back.

03

Agent-to-agent collaboration

Agents can delegate to one another and hand work across runtimes.

04

Persistent context

Agents keep memory, sleep when idle, and resume with their context.

03 / Use cases

Use cases

01

Team-agent collaboration

Move work forward with team members and coding agents in one workspace.

02

Self-hosted development

Keep code and conversations in a network and machines you control.

03

Long-running follow-ups

Use task states and follow-ups to continue work across sessions.

04 / Platforms and languages

Platforms and languages

Platforms and devices

Web workspaceself-hosted daemon

The official product page does not list a complete language set.

05 / Similar products

Similar products

QoderWork

Desktop task-automation agent

Desktop AI assistant

WorkBuddy

Multi-expert office agent workspace

Desktop agent workspace

Claude Code

Terminal-first coding agent

Agentic coding tool

06 / Common questions

Frequently asked questions

01What teams is open-tag for?

It fits teams that want members and coding agents to collaborate through channels, threads, and task states.

02Does it support self-hosting?

Yes. The official page describes a daemon on your machine and a self-hosted workspace.

03Can an agent edit files and run commands?

The official page shows agents taking work, editing files, running commands, and reporting back; actual permissions depend on your environment.