# Per-capability terminal pane snapshots are runtime artifacts: each capability
# test writes "<id>.txt" here at its verification point, and capability-report
# reads them back into the manifest + dashboard (both of which ARE committed).
# The raw captures are non-deterministic (spinner frames, hostnames, ids), so we
# keep the directory but ignore the generated files.
*.txt
