diff --git a/.github/workflows/spinifex.yml b/.github/workflows/spinifex.yml
index c39ef612d..159a390f1 100644
--- a/.github/workflows/spinifex.yml
+++ b/.github/workflows/spinifex.yml
@@ -131,6 +131,19 @@ jobs:
       - name: Check Diff Coverage
         run: cd spinifex && scripts/diff-coverage.sh coverage.out --base HEAD~1
 
+      # Measure-only: how much of the build cache this job uploads is stale.
+      # Go trims at 5 days, so entries below that age band are dead weight
+      # carried in every restore and save. Reports, deletes nothing.
+      - name: Go build cache age census
+        run: |
+          C="$(go env GOCACHE)"
+          echo "total: $(du -sh "$C" | cut -f1) across $(find "$C" -type f | wc -l) files"
+          for d in 1 2 3 5; do
+            find "$C" -type f -mtime +"$d" -printf '%s\n' | awk -v d="$d" '
+              {s += $1; n++}
+              END {printf "older than %sd: %.2f GB across %d files\n", d, s/1073741824, n}'
+          done
+
   # In-process integration tier (gateway.SetupRoutes over embedded NATS
   # JetStream). Its own job so it runs in parallel with the unit tiers rather
   # than adding to their wall time, and it still blocks a PR — unlike e2e.yml's