eywa

Persist and retrieve cross-session handoffs for Claude Code.

7|3|Updated Feb 8, 2026
One-click install
npx skills add https://github.com/buildoak/eywa-continuum --skill eywa
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: eywa
Source: https://github.com/buildoak/eywa-continuum/tree/main
Command: npx skills add https://github.com/buildoak/eywa-continuum --skill eywa

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Eywa provides cross-session memory for Claude Code, enabling persistence of handoffs and retrieval of past context across sessions, reducing cognitive overhead and re-explanation.

Core Features & Use Cases

  • Handoff creation at session end (eywa extract) and retrieval at session start (eywa get)
  • Batch indexing for historical sessions (eywa-batch) to bootstrap memory and improve recall
  • CLI-driven workflow with environment-driven model configuration (EYWA_CLAUDE_MODEL, EYWA_OPENROUTER_MODEL)

Quick Start

Call eywa get at session start to retrieve past context, and eywa extract at session end to persist a new handoff.

Frequently Asked Questions about eywa

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

FAQPage Schema
How do I keep Claude Code context persistent across different sessions?

To keep context persistent across sessions, use a cross-session memory CLI to extract handoffs at session end and retrieve them at session start. This reduces cognitive overhead by eliminating the need to re-explain past work.

What is the best way to retrieve historical project context when starting a new coding session?

The best way to retrieve historical context is by running a retrieval command at session start. This fetches indexed past handoffs and runtime context, allowing you to seamlessly resume ongoing work across projects.

Can I batch index previous Claude Code sessions to bootstrap cross-session memory?

Yes, you can batch index historical sessions to bootstrap memory and improve recall. This batch processing converts past sessions into markdown, extracts handoffs, and stores them for ongoing retrieval.

Do I need to configure specific AI models for Claude Code handoff extraction and retrieval?

You need to configure AI models for handoff extraction and retrieval using environment variables. You can define model behavior through environment-driven configuration to validate and process session context effectively.

How does cross-session memory handle markdown conversion and handoff storage?

Cross-session memory handles markdown conversion by detecting sessions and extracting handoffs. These handoffs are then stored and indexed, providing an end-to-end workflow for validating and retrieving context.

Why does Claude Code lose context and require re-explanation between sessions?

Claude Code loses context between sessions because it lacks built-in persistent memory. Without a cross-session memory tool to index handoffs, you must manually re-explain previous work each time you start a new session.