reflect

Generate structured reflections and maintain memory health in CLAUDE.md.

23|7|Updated Sep 8, 2025

Search Tags:#claude #self-reflection #reflection #ai-assistant #memory-management #engram #memory-health

Ready to run

npx skills add https://github.com/pvillega/claude-templates --skill reflect-pvillega

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: reflect
Source: https://github.com/pvillega/claude-templates/tree/main/plugins/ct/skills/reflect
Command: npx skills add https://github.com/pvillega/claude-templates --skill reflect-pvillega

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Self-reflection and memory hygiene for CLAUDE's working memory, enabling structured learning from conversations and maintaining engram health across CLAUDE.md.

Core Features & Use Cases

Generate structured reflections automatically after conversations to capture decisions, mistakes, and outcomes.
Maintain memory health by consolidating learnings in CLAUDE.md and keeping engrams clean and actionable.
Triggerable workflows via phrases like let's reflect, or commands such as /reflect, /reflect review, /reflect consolidate for different modes.

Quick Start

Invoke /reflect to generate a reflection and update CLAUDE.md as needed.

Frequently Asked Questions about reflect

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

FAQPage Schema

How does self-reflection work for AI assistant memory management?▼

Structured self-reflection for AI assistants generates four-category reflections after conversations to capture decisions, mistakes, and outcomes, maintaining engram hygiene across long-running sessions by updating CLAUDE.md and memory files.

How do I generate and review structured reflections for Claude?▼

You generate and review structured reflections by triggering specific workflows: use /reflect to generate reflections and update CLAUDE.md, /reflect review to check memory health, and /reflect consolidate to merge engrams across memory files.

Do I need dependencies to maintain engram hygiene in CLAUDE.md?▼

No external dependencies are required to maintain engram hygiene in CLAUDE.md. The self-reflection process operates natively by reading and writing structured updates directly to CLAUDE.md and associated memory files without additional packages.

When should I use memory consolidation commands in long-running AI sessions?▼

Use memory consolidation commands in long-running AI sessions when engrams become fragmented or outdated. Running /reflect consolidate after project work merges learnings, cleans memory files, and keeps CLAUDE.md actionable for future context.

Can I trigger memory health reviews automatically after project work?▼

Yes, you can trigger memory health reviews after project work by invoking /reflect review. This evaluates the state of engrams within CLAUDE.md and memory files to ensure long-running session context remains clean, structured, and actionable.