tstockham96/engram

Overview

Engram is a TypeScript library that provides a universal memory layer for AI agents. It enables agents to store, retrieve, and consolidate information across sessions.

Best for

Best for
TypeScript developers building lightweight memory systems for AI agents

Use cases

• Adding persistent memory to conversational AI agents
• Enabling context recall across multi-turn interactions
• Consolidating agent knowledge from multiple inputs

How to use

Install

npx engram-serve

Tools exposed

• engram_remember
• engram_recall
• engram_ask
• engram_briefing
• engram_consolidate
• engram_surface
• engram_alerts
• engram_audit
• engram_checkpoint
• engram_connect
• engram_forget
• engram_entities
• engram_stats
• engram_ingest
• engram_import_obsidian
• engram_import_claude_code
• engram_powered_by
• GEMINI_API_KEY
• ENGRAM_LLM_PROVIDER
• ENGRAM_LLM_API_KEY

Tested with

Claude Code, Cursor, ChatGPT

Notes

Engram is a TypeScript library that provides a universal memory layer for AI agents. It enables agents to store, retrieve, and consolidate information across sessions.

39 stars on GitHub. Last updated 2026-05-04. Licensed MIT.

Use cases

• Adding persistent memory to conversational AI agents
• Enabling context recall across multi-turn interactions
• Consolidating agent knowledge from multiple inputs

Pros

• Dedicated memory layer simplifies agent state management
• TypeScript provides type safety for integrations
• Open source with a permissive license

Cons

• Small community (39 stars) indicates early maturity
• No documented storage backends or persistence guarantees
• Limited documentation and examples available

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.