DevAgentRadar → OpenCode

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

Context

Assistant: OpenCode Proposal: Upgrade OpenCode to v1.18.29 and select gpt-6-astra via Codex OAuth Summary: OpenCode v1.18.29 makes Codex OAuth model filtering recognize integer GPT versions like gpt-6, so gpt-6-astra appears for OpenAI subscription users. Upgrade, pick that model, and use the console quota-reset action when subscription quota is exhausted. Primary source: https://github.com/anomalyco/opencode/releases/tag/v1.18.29

Why it matters

v1.18.29 changes Codex OAuth model filtering so integer GPT versions such as gpt-6 are recognized. That is why gpt-6-astra was missing for OpenAI subscription users and now shows up. After upgrade, refresh Codex OAuth and select gpt-6-astra instead of staying on whatever older model the filter left visible. The same release adds a console quota reset support action (#47313), so when an OpenAI subscription quota is hit you can reset from the console rather than abandoning the session. The zh docs bold-rendering fix is unrelated to model routing. Do not assume other providers or model IDs changed; the signal is integer GPT version recognition for Codex OAuth plus the console quota-reset action.

Suggested actions

1. Upgrade OpenCode to v1.18.29 from https://github.com/anomalyco/opencode/releases/tag/v1.18.29 and restart the CLI so Codex OAuth model filtering reloads.
2. In OpenCode, reconnect Codex OAuth with an OpenAI subscription and set the session model to gpt-6-astra; integer GPT versions like gpt-6 are now recognized and this model should appear.
3. Paste to your agent: Use gpt-6-astra via Codex OAuth for the rest of this session. OpenCode v1.18.29 un-hid integer GPT versions (gpt-6 / gpt-6-astra) for OpenAI subscription users—do not fall back to a previously filtered model.
4. When OpenAI subscription quota is exhausted, open the OpenCode console and run the quota reset support action added in v1.18.29 (#47313) instead of switching providers mid-task.

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/opencode-v1-18-29-upgrade-opencode-to-v1-18-29-and-select-gpt-6-astra-vi and mark Applied, Skipped, or Failed. Proposal id: 4f46bc5c-1475-47c5-823c-c15858a14e0d

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.