Set Codex model_provider to gcp and select GPT-6.1 Terra
Codex CLI rust-v0.165.0 adds a built-in Vertex AI provider via model_provider = "gcp", lists GPT-6.1 Terra in catalogs, filters the session picker by repository with directory grouping for linked worktrees, enables mid-conversation reasoning-effort changes by default, and defaults fresh profiles to memory v2.Why this loop
This release changes how you route models, resume work, and keep memory. A built-in Google Cloud Vertex AI provider is now selected with model_provider = "gcp", so GCP shops can drop custom provider wiring. GPT-6.1 Terra is in the model catalogs, so new sessions can target it without a local catalog patch. The TUI session picker gained a repository filter and the agent command center groups by directory, which is the intended way to navigate sessions across linked worktrees. Reasoning-effort changes during a conversation are on by default for supported models, so you no longer need a new thread to retune effort. Fresh profiles default to memory v2 and can import external-agent memories into that store. Treat model-not-found provider errors as terminal (retries are stopped and the original message is kept). Experimental attached thread forks exist on the app-server API only; do not build day-to-day workflow on them yet.
Proposed actions
- Set
model_provider = "gcp"in your Codex CLI config so this repo uses the built-in Google Cloud Vertex AI provider from rust-v0.165.0. - Select GPT-6.1 Terra from the Codex model catalog for new sessions now that rust-v0.165.0 added it.
- In the Codex TUI session picker, apply the repository filter and use agent command center directory grouping to jump between sessions in linked worktrees.
- Create a fresh Codex profile so memory v2 is on by default, then import existing external-agent memories into that v2 store.
- On supported models, change reasoning effort mid-conversation instead of starting a new thread; rust-v0.165.0 enables this by default.
Agent prompt
Paste into your agent or query via MCP (get_agent_prompt) — free, no extra AI costCodex / agent CLI task
DevAgentRadar → OpenAI Codex
Task: evaluate and optionally implement follow-ups from this coding-assistant release.
Context
Assistant: OpenAI Codex CLI Proposal: Set Codex model_provider to gcp and select GPT-6.1 Terra Summary: Codex CLI rust-v0.165.0 adds a built-in Vertex AI provider via model_provider = "gcp", lists GPT-6.1 Terra in catalogs, filters the session picker by repository with directory grouping for linked worktrees, enables mid-conversation reasoning-effort changes by default, and defaults fresh profiles to memory v2. Primary source: https://github.com/openai/codex/releases/tag/rust-v0.165.0
Why it matters
This release changes how you route models, resume work, and keep memory. A built-in Google Cloud Vertex AI provider is now selected with model_provider = "gcp", so GCP shops can drop custom provider wiring. GPT-6.1 Terra is in the model catalogs, so new sessions can target it without a local catalog patch. The TUI session picker gained a repository filter and the agent command center groups by directory, which is the intended way to navigate sessions across linked worktrees. Reasoning-effort changes during a conversation are on by default for supported models, so you no longer need a new thread to retune effort. Fresh profiles default to memory v2 and can import external-agent memories into that store. Treat model-not-found provider errors as terminal (retries are stopped and the original message is kept). Experimental attached thread forks exist on the app-server API only; do not build day-to-day workflow on them yet.
Suggested actions
- Set
model_provider = "gcp"in your Codex CLI config so this repo uses the built-in Google Cloud Vertex AI provider from rust-v0.165.0. - Select GPT-6.1 Terra from the Codex model catalog for new sessions now that rust-v0.165.0 added it.
- In the Codex TUI session picker, apply the repository filter and use agent command center directory grouping to jump between sessions in linked worktrees.
- Create a fresh Codex profile so memory v2 is on by default, then import existing external-agent memories into that v2 store.
- On supported models, change reasoning effort mid-conversation instead of starting a new thread; rust-v0.165.0 enables this by default.
Config surfaces this release may change
- hooks — check your repo before applying
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/openai-codex-rust-v0-165-0-set-codex-model-provider-to-gcp-and-select-gp and mark Applied, Skipped, or Failed. Proposal id: 0b238068-53f5-4d3c-965b-72d06723f370
Deliverables:
- Short impact assessment
- Optional patch plan (files + steps)
- Do not invent APIs not in the source notes
Your loop
This browser · no sign-in · not shared as “you”After you run the prompt
Only you can mark this. Agents cannot write your loop.
Your decision stays on this device. A public tally appears after a few votes.