Install codex
codex update && codex app-server daemon updateBashcurl -fsSL https://chatgpt.com/codex/install.sh | shBashcode ~/.codex/config.tomlBashnano ~/.codex/config.tomlBashmodel = "gpt-6-astra"
model_reasoning_effort = "xhigh"
service_tier = "fast"
approvals_reviewer = "auto_review"
approval_policy = "on-request"
sandbox_mode = "workspace-write"
# model_context_window = 372000
# model_auto_compact_token_limit = 334800
[tui]
screen_reader_detection_done = true
status_line = ["model-with-reasoning", "current-dir", "thread-name", "context-used", "weekly-limit", "context-window-size", "used-tokens", "thread-title"]
status_line_use_colors = true
mouse_scroll_speed = 3.0
TOMLsudo apt install ripgrep socat bubblewrap socat -yBashplaywright: unavailable
puppeteer: unavailable
jsdom: unavailable
codex app-server daemon startBashcodex app-server daemon restartBashcodex app-server daemon bootstrapBashcodex app-server daemon updateBashcodex app-server daemon stopBashpgrep -af codex
echo "$(ps afx)" | grep codexBashWrite git title and description for staged changes (don't have line breaks in the points)Markdownreview both the staged/unstaged changes along with the code in general, find as many coding standard best practices, efficiency duplication, over complication, unnecessarily complex codex issues etc you can find, List them allMarkdowncodex --model gpt-6-astra \
-c model_reasoning_effort=xhigh -c service_tier=fast \
--config model_context_window=372000 --config model_auto_compact_token_limit=334800 \
-c approvals_reviewer=auto_review -c sandbox_mode=workspace-write
Bash
Leave a Reply