reflect:ingest

You can index memories from multiple AI coding agents into a unified knowledge base by running the reflect:ingest command, which harvests and archives memory files from Claude, Codex, Copilot, and Gemini. This automatically captures and dual-indexes your project data into GraphRAG and QMD formats for retrieval.

The best way to archive AI memory files is to use a global knowledge indexer that captures and dual-indexes memories into a GraphRAG and QMD knowledge base. This archives project-specific notes and agent memories automatically, ensuring past learnings are preserved for future retrieval.

Yes, you need the reflect-kb tool and access to your specified memory sources to use this indexing Skill. It serves as the required dependency for archiving and indexing memories from your various AI coding tools into the centralized knowledge base.

Yes, you can consolidate memories from Claude, Copilot, and Gemini into a single searchable archive. The indexer harvests memories from these multiple tools and projects, indexing them into a unified GraphRAG and QMD knowledge base for comprehensive capture and retrieval.

GraphRAG and QMD indexing should be used for AI coding agent memory to maintain a comprehensive, up-to-date knowledge base that makes recalling past learnings efficient. This dual-indexing approach prevents you from making the same coding mistakes again across different projects.

To start the knowledge indexing process for your AI coding projects, simply run the reflect:ingest command. This triggers the automatic harvesting, archiving, and dual-indexing of memories from your connected tools into the GraphRAG and QMD knowledge base.