Create and append a <=300 Chinese-character daily summary into LogSeq journals...
Summarize a LogSeq daily journal (today by default, or a specified log) and append a concise <=300 Chinese-character paragraph back into that day’s journal page.
list_pages (include journals) and find a page whose title starts with YYYY-MM-DD. LogSeq journal titles may be YYYY-MM-DD DayName (English), YYYY-MM-DD 星期X (Chinese), or just YYYY-MM-DD.[journal], then the one with the longest non-empty content (via get_page_content), then the English weekday variant.search to confirm. If still missing, ask the user for the exact page name before creating a new page.get_page_content to fetch the page content.create_page or update_page) and note it was empty before summarizing.scripts/codex_usage_stats.py.python3 /Users/wushitao/.codex/skills/logseq-daily-summary/scripts/codex_usage_stats.py --date YYYY-MM-DD.~/.codex/sessions/YYYY/MM/DD/*.jsonl, aggregates the last token_count.total_token_usage per session file, and extracts the local time range for the day (min/max timestamps).🤖 Codex CLI 消耗(自动统计)
⏱️ 时间段: YYYY-MM-DD HH:MM:SS–HH:MM:SS (+0800)
🧾 用量: N tokens
🧠 Token: 输入 A / 输出 B / 缓存输入 C / 推理 D
update_page to append the usage block (if not already present), then use a second update_page call to append the summary block 今日总结: ... (no leading "- ").Use MCP LogSeq tools:
search or list_pages to locate the journal pageget_page_content to readupdate_page to append the summarycreate_page only if necessary