Auto-notify on long Bash commands; homey-alert local fallback; nextcloud mount fix
- homey-alert: add a global PreToolUse hook that transparently times every Bash command Claude runs and fires a notification once a command crosses the configured threshold (default 5 min), without needing the skill to be invoked manually. Also makes homey-alert always fire a local notify-send alongside the ntfy push, since the ntfy server isn't always reachable. - nextcloud: fix sleep crash-loop under the restricted systemd PATH (bin/sleep -> coreutils store path). - azos-core bump (adds ncdu to the base suite).
This commit is contained in:
@@ -8,6 +8,23 @@ When running (or suggesting) a command that will take more than ~30 seconds —
|
||||
builds, deploys, downloads, test suites — append a notification automatically. Don't wait to
|
||||
be asked.
|
||||
|
||||
## Background / async tasks (Claude Code specifically)
|
||||
|
||||
If you (the agent) are running a command via Bash that could take longer than
|
||||
~10 minutes — large rclone/rsync transfers, builds, training runs — and it
|
||||
might run in the background (either because you set `run_in_background`, or
|
||||
because the harness auto-backgrounds it after its timeout), chain the
|
||||
`homey-alert` call **into that same command string from the start**. Don't
|
||||
wait until you notice it's taking a while and then think about notifying —
|
||||
by then the command is already running without it, and the harness's own
|
||||
"task finished" signal only reaches this session, not Aner's phone. Chaining
|
||||
upfront is the only way to get a real ntfy push the moment it actually
|
||||
finishes:
|
||||
|
||||
```bash
|
||||
rclone copy ... ; homey-alert -t "Backup" -p default -T "$([ $? -eq 0 ] && echo white_check_mark || echo x)" "rclone copy finished (exit $?)"
|
||||
```
|
||||
|
||||
## Local commands
|
||||
|
||||
Append after the command so it fires on success only (`&&`) or always (`;`):
|
||||
|
||||
Reference in New Issue
Block a user