migrate-memory

Migrate agent memory stores between flat files and scaffolded directory trees.

Updated Jul 2, 2026
One-click install
npx skills add https://github.com/tharunramagiri/bookoraaagent --skill migrate-memory-tharunramagiri
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: migrate-memory
Source: https://github.com/tharunramagiri/bookoraaagent/tree/main/.claude/skills/migrate-memory
Command: npx skills add https://github.com/tharunramagiri/bookoraaagent --skill migrate-memory-tharunramagiri

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill solves the issue of data loss when switching AI providers for an agent group, ensuring that durable facts, user preferences, and seed instructions are preserved during the transition.

Core Features & Use Cases

  • Cross-Shape Migration: Automatically translates memory between flat file structures (CLAUDE.local.md) and scaffold tree structures (memory/ directory).
  • Intelligent Distillation: Filters out conversational residue while retaining critical identity and behavioral instructions.
  • Use Case: When migrating an agent from a Claude-based provider to a Codex-based provider, this skill ensures the agent retains its personality and historical context without manual copy-pasting.

Quick Start

Ask the agent to migrate memory for the specified group after you have updated its provider configuration.

Frequently Asked Questions about migrate-memory

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

FAQPage Schema
How do I migrate agent memory when switching AI providers?

Agent memory migration works by reading source memory files, filtering out conversational residue, and preserving critical identity instructions and user preferences when writing to the target memory store during a provider switch.

Can I transfer memory from a flat file structure to a scaffolded directory tree?

The migration process requires access to the group workspace to read source memory files and write to the target memory store after updating the provider configuration for the specified agent group.

How does intelligent distillation handle conversational residue during memory transfer?

No, this Skill operates without external dependencies, relying solely on workspace access to read the source memory files and write the distilled identity instructions to the new provider-specific format.

What is the best way to preserve agent identity and historical context across different providers?

This Skill prevents data loss by distilling durable facts, user preferences, and seed instructions from the source provider, ensuring they are preserved and written to the target memory store.

Do I need to manually copy-paste memory files when switching from Claude to Codex?

The Skill handles provider transitions by reading the original memory files, automatically translating the data shape, and writing the distilled memory to the target provider format.

Why does my agent lose its behavioral instructions after a provider switch?

To resolve memory loss, update the provider configuration and ask the agent to migrate memory for the specified group, which reads the workspace files and writes preserved identity instructions to the target store.