| name | test-setup |
| model | haiku |
| description | Dry-run trace of setup.sh execution |
| disable-model-invocation | false |
| argument-hint | [--terminal|--godot|component] |
| allowed-tools | Read |
Trace setup.sh execution without running it — dry-run analysis.
If $ARGUMENTS provided (e.g., --terminal, --godot), focus on that component only.
-
Read the scripts:
setup.sh (orchestrator)
scripts/common.sh (shared utilities)
- Each component script that would execute
-
Trace execution path: Determine:
- Which component scripts run, in what order
- Files copied/created/modified (source -> destination)
- Packages installed
- Commands executed
-
Check for issues:
- Missing source files (every
cp/ln target exists in assets/?)
- Permission needs (sudo operations?)
- Existing file conflicts (destination already exists?)
- Package availability (
dnf info <pkg> without installing)
- Idempotency issues (
mkdir without -p, ln without -f)
-
Report per component:
- What it installs/configures
- Source -> Destination file mappings
- Dependencies satisfied?
- Risk level (safe / caution / risky)
-
Overall GO / NO-GO recommendation.