dakdevs

OS Assistant: Skill+Bash Architecture vs Native Tools — Full Comparison

Date: 2026-04-03 Branch under test: autoresearch/apr03 (skill+bash) vs main (native tools) Eval harness: os-qa automated scenario runner (84 scenarios)

---

Table of Contents

dakdevs

QA Comparison: Main vs Skills Architecture

Side-by-side comparison of the native tool architecture (main) vs the skill+bash architecture across 85+ QA scenarios. Both runs use the same Qwen 3.5 397B model on the same hardware, same NATS infrastructure, same QA evaluator. Only the os-assistant binary and scenario acceptance criteria differ.

Skills scored 48 GREAT vs main's 44 with the same BAD count (23 vs 23). Token usage was comparable (10.3M vs 10.3M) with fewer LLM calls (4,869 vs 5,649). Median time to first audio was 1.5s (skills) vs 1.7s (main).

Grade Summary

	Main	Skills

dakdevs

#!/usr/bin/env bash
# install-xrdp.sh — set up remote desktop (xrdp + Xfce) on a remote Ubuntu box
#
# Run this on your Mac. It will:
#   1. Check for / install the Mac RDP client (Windows App) via Homebrew or App Store
#   2. Prompt for the SSH host and Linux user to configure
#   3. scp itself over, run via sudo, clean up after
#   4. Install xrdp + xorgxrdp + xfce4 on the remote
#   5. Configure ~/.xsession to launch Xfce with an isolated dbus bus
#      (so it coexists with any GNOME session on the physical monitor)

dakdevs

#!/usr/bin/env bash
# Setup an Ubuntu machine to be SSH-able from dakdevs' Mac.
# Idempotent: safe to re-run.
#
# Usage (on the Ubuntu machine):
#   curl -fsSL <gist-raw-url> | bash

set -euo pipefail

PUBKEY="ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIEqJZkbO069wsfHTuRDMPLNe8SWozP5Gf/cpTgxT3XI1 dakdevs-mac"