rem-sleep
Consolidate agent sessions into linked, low-noise long-term memory
- Stars
- 0
- Forks
- 0
- Updated
- Updated Jun 24, 2026
Consolidate agent sessions into linked, low-noise long-term memory
REM Sleep processes an agent’s session trace into atomic, retrievable memories, links related lessons, prunes noise, and separates useful guidance from the emotional charge of an event. Its default store is a file-based index with one fact per file and frontmatter, but the stages can adapt to another format after asking once. The skill is intended for cross-session learning and works with Claude Code, Codex, and Cursor without requiring a database.
Resource types
Use cases
Platforms
Capabilities
Turn project files into tiered context packs and an Obsidian knowledge base.
Keep product strategy and business constraints in the repository for AI agents.
Public GitHub facts last synced Jul 10, 2026.
Preserve project decisions and constraints across sessions and agent handoffs.
Give AI agents local, searchable memory backed by Markdown and SQLite.