Linked Terminals

Delegate work to an AI terminal you already have open, and get the answer back.

You have Claude reading a spec in one pane and Codex holding the whole repo in another. Handing work between them used to mean copying an answer out of one terminal and pasting it into the other.

Linked Terminals removes that step. Mention a terminal that is already open and your prompt is delivered into that live agent — not into a new one that knows nothing about either conversation.

The mention picker's Terminals tab listing open AI terminals grouped by project

Mention a terminal

In Agent Input, type @. The mention picker opens with four tabs: Files, Agents, Templates, and Terminals.

The Terminals tab lists every open AI terminal, grouped by project and searchable by terminal name or project name.

  • Tab — switch between tabs
  • ↑↓ — move through the list
  • ↵ — insert the mention at the cursor

The mention lands inline and highlighted, so the prompt reads like a sentence that names its own collaborators:

use codex full to review the docs/proxy.md then claude fable update the fixes and codex full review again until both APPROVE.

The reply edge

A link is directional. Mentioning codex full lets you send work to Codex. It does not give Codex a way to send anything back.

To open the return channel, mention this terminal in the same message. The picker marks it with a reply edge hint.

With the reply edge in place:

  • The answer arrives in your terminal instead of being printed in the peer's.
  • Every delivered message carries who sent it and the exact command to answer with.
  • When the reply lands, the original request is marked answered instead of sitting on delivered forever.

Without it, the peer is told it has no channel back. It will do the work, print the result locally, and stop. Those delegations show up as stuck in Mission Control.

Agent Input with two open terminals mentioned inline and the orchestration skill detected

Team, Swarm, or Off

When you name collaborators, 1DevTool detects it and offers a topology:

  • Team — named collaborators with a host that integrates the results.
  • Swarm — peers that vote, with no master. Open decisions are visible while agents vote.
  • Off — deliver the prompts, no orchestration wrapper.

Routing roles is a separate toggle. Off means only the agents you mentioned participate, and table roles are not applied.

Links can cross projects

The terminal you mention does not have to be in the current project. An agent in one repo can hand work to an agent in another.

The link is owned by the sender's project but stays visible from both ends, so the receiving project never shows an agent being driven by an invisible peer.

What happens when you send

  • A project running live orchestration switches to the 2×2 grid so you can watch the participants. If you change the layout yourself, your choice always wins.
  • Mission Control appears above the terminal grid.
  • Edges on the Team Map pulse green as each message is delivered.

Which agents can receive a delegation

Claude Code, Codex, Gemini, Antigravity, Amp, OpenCode, Cline, Qwen, Grok, and Hermes.

If an agent you mention has no terminal open, 1DevTool starts one. The point of Linked Terminals is that when the terminal already exists, it is reused instead of a duplicate spawning beside it.

From the command line

The agent CLI covers the same ground, so an agent can orchestrate without touching the UI:

bash
# Who am I, and what am I linked to?
1devtool-agent whoami

# Send work to a linked terminal and wait for the answer
1devtool-agent link send --to=<terminalId> --prompt-stdin --wait

# Answer a specific request
1devtool-agent link send --to=<terminalId> --prompt-stdin --reply-to=<messageId>

# One board of everything you sent and everything waiting on you
1devtool-agent link status

Run whoami first — it reports your terminal id and your current links, which is where the --to= value comes from.

Next