@YuLin807: This is a very valuable official Codex self-improvement prompt: Recommended use: "Review my recent work over the past 30 days (or all available history if shorter), and identify repetitive manual workflows worth packaging. Use available evidence in the following order: - Most...

This is a very valuable official Codex self-improvement prompt: Recommended use: "Review my recent work over the past 30 days (or all available history if shorter), and identify repetitive manual workflows worth packaging. Use available evidence in the following order: - Recent Codex session and task summaries. - Codex memory and rollout summaries to find patterns that repeat across sessions. - If enabled, chronicles to discover repetitive work outside Codex. Only use chronicles for discovery; confirm important details in relevant source systems whenever possible. - Existing skills, custom agents, and automations to reuse or extend what already exists, rather than recreating it. Broadly look for work that is repetitive, time-consuming, error-prone, context-heavy, or benefits from a consistent process. Include workflows across coding, research, writing, planning, communication, operations, analysis, and personal management. Only take action on a candidate when it meets the following criteria: - Has occurred at least twice, or is clearly likely to repeat with high cost of repetition; - Has stable inputs, repeatable procedures, and clear outputs or stopping conditions; - Would substantially improve speed, quality, consistency, or reliability; - Is not already adequately covered. Choose the smallest appropriate form: - Skill: reusable workflow or playbook. - Custom sub-agent: a well-defined expert role or investigative task suitable for delegation. - Automation: scheduled or recurring checks, reports, reminders, or monitoring. - Skip: work that is too one-off, vague, sensitive, or lacking sufficient evidence to package. First generate a concise shortlist including: - Repetitive workflow - Supporting evidence and dates - Frequency/confidence - Recommended form: skill, sub-agent, automation, extend existing, or skip - Why it is or is not worth creating Then only create high-confidence missing items. Keep them narrow, practical, source-aware, and easily verifiable. Do not create speculative, overlapping, or overly broad assets. Finally, end with: - What you created or extended - What you deliberately skipped - What needs more evidence before packaging"