OptMem provides a permanent memory system for AI agents via a 426-token prompt and a script, enabling memory recording and retrieval with a small CLI executable. It ships with a simple installer, configuration options, and memo commands to wake, note, recall, and manage memories.
What it is
Permanent memory for AI agents. A 426-token prompt, a script, plug and play.
How it works
Installs a Python-based memory tool that stores memories under ~/.optmem/memory with a memo executable that supports wake, note, nap, recall, zoom, and forget. The installer prints a Memory block to paste into AGENTS.md. The tool maintains a memory log and a tree-based summaries cache.
Getting started
Install using the provided command:
curl -fsSL https://raw.githubusercontent.com/VictorTaelin/OptMem/main/install.sh | sh
Paste the Memory block at the top of your agent's AGENTS.md (or CLAUDE.md), then run the same line again to update. The tool lands at ~/.optmem/memo; put ~/.optmem on PATH to type memo.
Usage commands include:
memo wake
memo note "..."
memo nap
memo recall <regex>
memo zoom <lo>-<hi>
memo forget <lo>-<hi>
Getting started (config)
memo config # show the sizes
memo config WAKE_LINES=300 # how many lines wake prints (208 ≈ 16k tokens)
memo config WAKE_LINES= # back to the default
WAKE_LINES controls the wake printing budget; it does not affect storage. Records are fixed width and lookup is one seek.
Recent releases
RELEASES (latest 0):
- none
Traction
stars_1d: 134 stars: 773 open_issues: 0
Caveats
license: none listed created: 2026-07-25 last_push: 2026-07-27 language: Python