Capture session learnings into memory/lessons.md and memory/patterns.json.
npx skills add https://github.com/ngocsangyem/MeowKit --skill meow-memory
Please help me install this Agent Skill. Skill: meow:memory Source: https://github.com/ngocsangyem/MeowKit/tree/main/.claude/skills/meow%3Amemory Command: npx skills add https://github.com/ngocsangyem/MeowKit --skill meow-memory
Capture and persist session learnings to enable cross-session memory for MeowKit workflows.
End-of-session save enables memory reuse for future analyses and improvements.
High-intent search queries and answers about installing and using this skill.
You can persist session learnings by capturing decisions, patterns, and failures at the end of each run. The system stores these entries in memory files to enable cross-session memory and surface insights for future workflows.
Tracking recurring patterns in session memory involves recording insights and updating their frequency and lastSeen values. This pattern consolidation surfaces recurring insights automatically for future review and analysis.
Cross-session memory for cost tracking works by capturing session data at the end of each run and storing it in structured files. This provides hooks for budgeting and allows you to review cost patterns across previous sessions.
Yes, you can store session decisions and failures using markdown and JSON files. The system specifically writes entries to lessons.md for text records and patterns.json for structured frequency tracking and consolidation.
You need to save session memory at the end of each workflow run. Capturing this data at the end of a session ensures that decisions and failures are available to surface as insights for future analyses and improvements.
No external dependencies are required to enable cross-session memory persistence. The system operates independently to capture session learnings and store them directly in local memory files for future reference and cost tracking.
calcosmicGenerate a SESSION_REPORT.md from Git logs, file changes, and phase progress.
phuetzSearch and analyze Code Buddy session logs in JSON files.
usuroborCapture single transferable insights with trigger context for agent learning.
abzhawCoordinate in-context, working, long-term, and episodic memory systems for LLM agents.
