Problem
If you send a prompt while the agent is mid-turn, Helm gives no feedback about whether it will be queued, steer the current turn, or run after. In the demo, a sent prompt is just appended and "Queued your instruction" comes back — but the UI never communicated that up front.
Suggested fix
Surface a "will run after the current step / queued" hint on the composer when a turn is active, and ideally a small pending-queue list (à la VS Code Copilot Chat's queue + drag-to-reorder).
Reference
https://code.visualstudio.com/docs/chat/chat-overview ("Add to Queue", "Steer with Message").
Filed from a UI research sweep of the Helm mobile app (landing + live demo session) using Playwright + full source review, cross-referenced against Claude, ChatGPT, GitHub Copilot Chat, and Codex/Claude Code UIs.
Problem
If you send a prompt while the agent is mid-turn, Helm gives no feedback about whether it will be queued, steer the current turn, or run after. In the demo, a sent prompt is just appended and "Queued your instruction" comes back — but the UI never communicated that up front.
Suggested fix
Surface a "will run after the current step / queued" hint on the composer when a turn is active, and ideally a small pending-queue list (à la VS Code Copilot Chat's queue + drag-to-reorder).
Reference
https://code.visualstudio.com/docs/chat/chat-overview ("Add to Queue", "Steer with Message").
Filed from a UI research sweep of the Helm mobile app (landing + live demo session) using Playwright + full source review, cross-referenced against Claude, ChatGPT, GitHub Copilot Chat, and Codex/Claude Code UIs.