What problem does it solve? Diagnosing failing NAS disks requires manually SSHing into the device, running many smartctl and mdadm commands, and interpreting raw SMART attributes—a slow, error-prone process where a missed flag (like -d sat for expansion-card disks) produces misleading results. ## Core Features & Use Cases - Remote Read-Only Collection: Connects via SSH (paramiko) to Synology and similar NAS devices to gather filesystem usage, RAID array status (/proc/mdstat, mdadm --detail), block devices, controller info, and dmesg ATA/I/O errors. - Deep SMART Diagnostics: Reads full SMART attributes, error logs, and self-test logs with the critical -d sat flag for disks behind expansion cards (e.g., ASMedia ASM1166), avoiding false "no SMART capability" reports. - Repairability Assessment & Visual Report: Applies a decision tree over key SMART attributes (IDs 5, 187, 188, 197, 198, 199, 9) to judge whether a failed disk is recoverable, then renders a single-file HTML report with status cards, slot grid, log-scale SMART charts, and power-on-hours comparison. - Use Case: Your Synology RAID array shows degraded status. Run the skill to locate the failed physical disk, determine from SMART raw values whether it can be rescan-recovered or needs replacement, and deliver a shareable HTML report with tiered action recommendations. ## Quick Start Ask the assistant to diagnose the disks on your NAS at a given IP with your SSH credentials and produce a health report.
