DevAgentRadar → GitHub Copilot

You are helping me adopt a real coding-assistant change. Work only from the facts below. Do not invent features.

Context

Assistant: Kiro Proposal: Enable Kiro Start fullscreen and route repo-scale work to GPT-5.6 1M context Summary: Kiro 2.22.1 adds /fullscreen chat for TUI and Lite (V2/V3), a redesigned V3 /sessions dashboard, and 1M-token GPT-5.6 Sol/Terra/Luna windows. Turn on Start fullscreen, resume from the new session list, and send repo-scale work in one request—rates double above 272K tokens. Primary source: https://kiro.dev/changelog/#v2-22-1

Why it matters

Kiro v2.22.1 changes how long interactive work should run. Fullscreen chat (/fullscreen) moves the current TUI or Lite conversation (V2 and V3) to a dedicated terminal screen with its own scrollable view, mouse/touchpad/keyboard scroll, and click-drag copy, independent of shell history; /fullscreen again returns inline. Persist that via /settings → Display → Start fullscreen. The V3 /sessions dashboard is now a flat last-use list with search, filter, sort, and grouping split from the list, Tab between controls and results, sort by last use, session name, or message count, and a remembered selection—so resume the existing long session instead of spawning another. GPT-5.6 Sol, Terra, and Luna now support a 1M-token window (up from 272K) in Kiro IDE, CLI, and Web, so repository-scale refactors, long documents, and multi-turn agent histories can stay in one request. Credits follow a two-tier model: ≤272K at Sol 4.4x / Terra 2.2x / Luna 1.1x; above 272K those rates double. Prefer the short tier unless the extra context is required; 1M context is rolling out gradually.

Suggested actions

1. In a Kiro interactive TUI or Lite session, type /fullscreen to move this conversation to a dedicated terminal screen with independent scrolling and click-drag copy; type /fullscreen again to return inline.
2. In Kiro, type /settings, choose Display, and turn on Start fullscreen so future interactive TUI sessions launch in fullscreen.
3. In Kiro V3, type /sessions, press Tab to move between the separated search/filter/sort/grouping controls and the session list, sort by last use, session name, or message count, and resume the remembered selection instead of starting a new session.
4. For repository-scale refactors, end-to-end debugging, or long-horizon agent work in Kiro IDE, CLI, or Web, select GPT-5.6 Sol, Terra, or Luna so the full codebase and multi-turn history fit in one 1M-token request without chunking.
5. Before sending a GPT-5.6 Sol/Terra/Luna request, keep context at or below 272K tokens to bill at Sol 4.4x / Terra 2.2x / Luna 1.1x; only exceed 272K when 1M context is required, because those multipliers double above that tier.

After you finish

Do not report this as applied to DevAgentRadar. You cannot write the visitor's loop.

Tell the human: open https://devagentradar.com/proposals/kiro-v2-22-1-enable-kiro-start-fullscreen-and-route-repo-scale-work-to-g and mark Applied, Skipped, or Failed. Proposal id: 1cfa2170-50f7-443a-80cc-b45e1d25bbac

Your job

1. Restate the change in one sentence.
2. Propose a minimal plan for my repo (or a throwaway pilot).
3. Implement only what I approve; prefer small diffs and tests.
4. Call out risks (permissions, breaking APIs, cost).

Start by confirming you understood the proposal.