Agentmemory Config

This is the reference card you'll want when you're toggling features or troubleshooting why agentmemory isn't behaving the way you expect. It covers the environment variables that control token-spending features like auto-compress and context injection, both of which ship disabled because they scale with how often you call tools. The port layout is clearly explained (REST at 3111, then streams, viewer, and engine offset from there), and you'll learn that agentmemory works fine without any API key by falling back to BM25 and local embeddings. If you're setting up auth, changing the tool visibility level, or relocating the whole port block with instance flags, this is where you start.

skills.sh installs: 11.5K
GitHub stars: 30.2K
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add rohitg00/agentmemory --skill agentmemory-config --agent claude-code

Installs into .claude/skills of the current project.

Files

agentmemory reads configuration from the environment and from ~/.agentmemory/.env (one KEY=value per line, no export prefix). Restart the server after changing it.