context-graph-compressor

To compress chat history into a JSON context graph, instruct the assistant to "compress this chat," which converts conversations into a portable graph preserving decisions and context for cross-session continuity across different LLMs.

A context graph is a portable JSON structure that maps conversations using typed nodes (F, D, P, G, C, A, X), importance levels, and relationships to preserve decision history and context when handing off to a new chat or different LLM.

You can save tokens during chat context transfer by using the compact handoff mode, which encodes conversations into a token-efficient JSON context graph while retaining critical decision history for the new session.

Yes, the context graph handoff supports cross-session continuity across different LLMs and platforms by outputting a downloadable, portable JSON file that any compatible AI system can ingest.

The best way to archive AI chat history is using the readable mode of the context graph compressor, which structures conversation decisions and context into a human-readable JSON format suitable for long-term archiving.

The compressed JSON context graph enforces a strict schema utilizing specific node types (F, D, P, G, C, A, X), importance levels, statuses, and relationships to accurately map the conversation history.