From e9561b90d692dfec2fc3e7ae85a0d1cdbbf766ec Mon Sep 17 00:00:00 2001 From: My first Letta Agent Date: Wed, 12 Aug 2026 20:15:25 -0400 Subject: [PATCH] Add external memory links for discoverability --- system/about_user.md | 3 +++ 1 file changed, 3 insertions(+) diff --git a/system/about_user.md b/system/about_user.md index 4cb18e0..946af4f 100644 --- a/system/about_user.md +++ b/system/about_user.md @@ -18,6 +18,9 @@ description: Key facts about the user. - Hates being told what to do — no "go do X" or "you should Y" directives. Just give info and let him decide. - Technical interests: DevOps, self-hosting, infrastructure, cloud services, CLI tools - Actively researching AI agent behavior — context management limitations, "bad-tripping" (agents spiraling/can't self-correct despite memory), persistent memory drift. Applies LLMs to coding (custom Emacs harness with controlled context, ~2024+) +- Infrastructure details, server configs, and DevOps patterns: [[reference/tech_context]] +- Active project: Zulip scheduler bot [[reference/projects/zulip-scheduler-bot]] +- Job applications: [[reference/tailscale-application]] and [[reference/keycard-labs-2026-08]] - Uses Arch Linux on personal machine; deep OS internals knowledge from extensive personal debugging experience (days cumulatively spent debugging Linux internals) - Fast learner — develops deep expertise personally that can translate quickly to professional context - Package manager preference: Homebrew (on macOS) -- knotmirror 0.0.1