mem-weekly

Analyze weekly L1 context records and generate L2 behavior candidates.

1|Updated May 12, 2026
One-click install
npx skills add https://github.com/cocyuhao/my-ai-skills-library --skill mem-weekly-cocyuhao
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: mem-weekly
Source: https://github.com/cocyuhao/my-ai-skills-library/tree/main/mem-weekly
Command: npx skills add https://github.com/cocyuhao/my-ai-skills-library --skill mem-weekly-cocyuhao

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

本 Skill 提供一个自动化的周复盘工作流,用于从本周的情境记录中识别重复模式并将其提炼到 L2 行动层,帮助用户高效地总结与改进行为。

Core Features & Use Cases

  • 自动化周复盘流程:在用户触发时,按事前定义的步骤自动完成从 L1 记录读取到 L2 更新的全过程。
  • 重复模式识别与候选生成:统计本周出现 >=3 次的行为/偏好,生成待确认的候选清单。
  • 可控的人工确认:在输出候选后等待用户确认再将结果写入 L2,以防误判。 使用场景包括:用户说“周复盘”时、每周定期回顾时、以及从日常记录中提炼行为模式時。

Quick Start

直接在应用中触发“周复盘”,系统将按流程扫描、分析并给出候选清单以供确认。

Frequently Asked Questions about mem-weekly

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate weekly review and extract recurring behavior patterns from daily notes?

Automated weekly review scans your daily L1 context records to identify recurring patterns, generating actionable L2 behavior candidates. It uses a deterministic workflow with pre-scan steps, candidate generation, and structured updates to streamline retrospective analysis.

How does pattern detection work when deriving behavior patterns from L1 records?

Pattern detection identifies recurring behaviors or preferences appearing three or more times within your weekly L1 context records. It generates a candidate list for user confirmation, ensuring accuracy before finalizing structured L2 behavior updates.

Can I require manual confirmation before updating L2 behavior candidates during a retrospective?

Yes, the weekly review workflow requires user confirmation before finalizing changes to L2 behavior candidates. This prevents automated misjudgments and ensures deterministic updates align with your actual observed patterns.

What is the best way to perform a periodic retrospective using Obsidian daily records?

Periodic retrospectives work by reading your daily L1 context records, statistically identifying recurring patterns, and outputting a structured candidate list. You then confirm these candidates to update your L2 behavior layer.

Does the weekly review workflow work without external dependencies?

Yes, the weekly review workflow operates independently without external dependencies. It uses a deterministic internal process to scan L1 records, generate behavior pattern candidates, and apply confirmed L2 updates.

When should I not use automated pattern extraction for weekly reviews?

Avoid automated pattern extraction when your daily L1 context records lack sufficient recurring behaviors, as the system requires patterns appearing three or more times to generate meaningful L2 behavior candidates.