RepoGo Logo
Documentation

Features

Voice Agent

Last updated: 2026-09-21

The Voice Agent lets you run RepoGo hands-free. Instead of typing, you talk: you can spin up new chats, send work to agents already running, check on what they're doing, and answer their questions — all by voice, in real time.

By default it is a global orchestrator, not a single chatbot: one voice session can coordinate chats across your numbered environments and workspaces. You refer to chats by their spoken handles ("Blue Otter, run the tests") or catalog numbers, and the agent routes your instructions to the exact target. You can also confine a session to one project — see below.

Project-scoped sessions#

Where you start voice decides its scope, and the scope is fixed for the whole session. A scoped session only sees that project's chats and summaries, new chats always land in it (there is no workspace to pick, so nothing can be started in the wrong repo), and history search only covers that project. A chip above the chat cards names the scoped project.

On iOS, the entry point decides:

  • From a project's chat list — scoped to that project.
  • From the drawer root, Siri, or the Action Button — global, even if a project happens to be on screen.

On the web dashboard, the session takes the workspace selected in the sidebar when you start it. With no workspace selected, it's global.

Changing scope means stopping and starting a new voice session from the project you want — a live session never silently re-scopes, so switching the sidebar workspace mid-session leaves the agent where it started.

What it does#

Ask which workspaces and chats are available, then refer to them by their spoken handles or numbers. You can:

  • Start new work. "Open a new chat called Auth Refactor and have Claude split the login view into smaller components." The agent creates the chat and dispatches your prompt.
  • Send work to existing chats. "Blue Otter, also add tests for that." The message is routed to that chat — sent immediately if it's idle, or queued behind the in-flight turn if it's busy.
  • Check progress. "What's Red Fox up to?" The agent summarizes the chat's current status, latest assistant message, the tools it has called, and the files it touched — in plain language.
  • Search your history. "How many times did I run a shell command today?" The agent runs a read-only query over your own chat history.
  • Approve or deny tool requests. When an agent pauses for permission ("can I run this command?"), you can say "approve" or "deny" — or "always approve" to stop being asked for that tool this session.
  • Answer agent questions. When an agent asks you a clarifying question, you can answer it by voice.

With GPT Live on iOS, voice can also list, open, focus, and close browser and terminal tabs in a workspace, navigate pages, read page snapshots, click, type, press keys, scroll, wait for page content, and run browser JavaScript. Ask “what tabs are open in repogo?” before referring to a tab by number. Browser and terminal numbers are separate and follow the grid order; they can change when tabs close or move.

GPT Live on iOS can also list and run your Actions. Ask “What actions are available?” and name the one to run.

Bring your own API key#

The Voice Agent uses your provider account — you bring your own API key. RepoGo never marks up usage; you pay the provider directly at their published rates.

Add your key under Settings → Integrations, in the Voice Agents section. Each provider has its own entry:

Keys are encrypted at rest. Choose a provider in the voice panel; RepoGo remembers your choice between launches.

Models and cost#

Choose a provider in the voice panel. See Supported Agents, Models & Voices for the available models.

Usage is billed by your provider. Check its current rates before a long session: OpenAI, Gemini, or xAI.