# Use speech, visuals, and marked regions as one task context

> 1 — Start recording and speak naturally. Do not try to make the first sentence perfect. The live waveform shows that audio is being received, and the small…

Canonical: https://optionos.app/guide/en/voice-context-workflow
Published: 2026-07-19
Source: optionos-product:r19

**1 — Start recording and speak naturally.** Do not try to make the first sentence perfect. The live waveform shows that audio is being received, and the small progress indicator shows that speech is becoming text in the background.

Release note visual: https://optionos.app/whats-new/en/

**2 — When you say “this part,” capture the screen and mark the region.** Choose Screen Capture from Quick Actions or use the capture shortcut. The rectangle enters the conversation as a clickable O1.R1-style reference, so you do not need to write a separate explanation list.

Release note visual: https://optionos.app/whats-new/en/

**3 — Check which app and visual the conversation belongs to.** The live flow shows the target app. Conversation opens plain speech, Events exposes the structure sent to the agent, and Context opens the combined human-readable explanation. Each captured-item row carries its source app, representation, and approximate magnitude.

**4 — You do not need to paste immediately.** If you start another recording, the previous unpasted recording waits. When you deliver, pending recordings merge into one session instead of making you rebuild context by hand.

**5 — Find an older conversation in Dictation Browser.** Search by duration, user tag, relationship, text, or a suggested filter. Select a word, pause, or waveform position in the chosen recording to play audio from there.

Release note visual: https://optionos.app/whats-new/en/

**6 — Follow the context tree through nested visuals.** The first tour moves from the assembled prompt into recording history, live Dictation, the hover preview, and Cockpit. Its other branch moves from the capture stack to the Codex delivery. Each screen opens from the context of the previous visual.

Release note visual: https://optionos.app/whats-new/en/

**7 — Watch the Cockpit card while the agent works.** The card shows the goal, terminal, first and last human and agent messages, subagent conversations, and file states. Copy only human messages, take editable context, fork, or use the optionOS Compact action when needed.

Release note visual: https://optionos.app/whats-new/en/