new·The score now tells you which way it movedA brain's exam only ever grows: its own material writes questions, and so does every question a real caller asked and did not get answered. The score is a percentage over that growing set, so a brain that learned more could post a smaller number — and this week three did. One of them answered two MORE questions than the week before and showed eighteen points less. Printed as a single percentage, that reads as decline to a reader and as punishment to anyone who contributes material.all news →
mozg.beta
Sign in

AI SDK · Providers · all subjects

codex-cli/options

3 notes, read out of this brain and free to use. Each one was extracted from a source and is re-checked against its exam.

Codex CLI model configuration options

The createCodexCli function accepts a configuration object with a defaultSettings property. Model settings that can be configured include: reasoningEffort ('none' | 'minimal' | 'low' | 'medium' | 'high' | 'xhigh'), approvalMode ('untrusted' | 'on-failure' | 'on-request' | 'never'), sandboxMode ('read-only' | 'workspace-write' | 'danger-full-access'), mcpServers (Record<string, McpServerConfig>), verbose (boolean), and logger (Logger | false).

Codex CLI per-model setting override

Settings configured in createCodexCli defaultSettings apply to every model. Individual models can override these settings by passing a settings object as the second argument to the model creation function, for example: codexCli('gpt-5.1-codex', { reasoningEffort: 'high' }).

Codex CLI reasoning configuration example

Reasoning can be configured per-model: const model = codexCli('gpt-5.1-codex-max', { reasoningEffort: 'high', reasoningSummary: 'detailed' }). The reasoningEffort option accepts 'none', 'minimal', 'low', 'medium', 'high', or 'xhigh'.

Give your agent this brain