Merge pull request #120 from dan-claude-bot/fix/wipe-sigpipe-shape

fix: drill/wipe.sh reads ufw into a capture, not into an early-exit reader

CHANGELOG.md

@@ -374,6 +374,31 @@ which records not just what changed but what each drill run proved.
   and *above* the first `incus` call, so the refusal costs no daemon and
   `test/cli.sh` drives it for real rather than grepping for it.
 
+- **`drill/wipe.sh` no longer carries #102's SIGPIPE shape, and the pin now
+  sweeps the class** (#107) — the file piped `ufw status` straight into
+  `grep -q "Status: active"`. `Status: active` is ufw's FIRST line, so the
+  reader matches, closes the pipe, ufw dies of SIGPIPE, and the pipeline
+  yields 141. This was **correct today and only by accident**: `wipe.sh` is
+  `set -u` with no `pipefail`, so the 141 was discarded and grep's 0 carried
+  the branch. It was also one line from wrong — adding `set -o pipefail` for
+  unrelated robustness, the kind of tweak that reads as an obvious
+  improvement, would have silently skipped every UFW removal on a host the
+  operator was told is wiped, with no error and no red X (measured on a shim:
+  5/5 runs took the wrong branch under `pipefail`, 3/3 the right one
+  without). Now it captures `ufw_status` once and matches with `[[ ]]`, and
+  the numbered-delete loop — whose condition was itself an early-exit reader,
+  plus an un-captured re-read to get the number — reads one capture per
+  iteration and breaks on absence. That is #106's pattern transplanted
+  verbatim from `host/teardown-host.sh`, so both files now read alike.
+  The `test/cli.sh` pin is **generalized from the single site to the class**:
+  it sweeps every `host/*.sh` and `drill/*.sh` for the racing shape and names
+  the offenders, so a new script in either directory inherits the pin instead
+  of being one more site to remember. Comment lines are stripped before
+  matching — each fix's own commentary quotes the racing shape to explain it,
+  and a prose-blind pin would fail on the very comment documenting why it
+  exists. (`drill/doctor.sh` was checked and needs nothing: it already reads
+  into `ufw_out` and is safe by construction, not by absent `pipefail`.)
+
 ## 0.8.0 — 2026-07-19
 
 ### Added

drill/wipe.sh

@@ -117,10 +117,39 @@ for t in "inet box" "bridge box" "inet claudebox" "bridge claudebox"; do
   # shellcheck disable=SC2086 # the table spec is two words by design
   sudo nft delete table $t >/dev/null 2>&1 && say "deleted nft table $t"
 done
-if command -v ufw >/dev/null && sudo ufw status 2>/dev/null | grep -q "Status: active"; then
+# Every ufw read is CAPTURED before it is matched, never piped into a reader
+# that exits on its first match (#102, #107).
+#
+# `ufw status | grep -q "Status: active"` returns the WRITER's exit: grep
+# matches on the first line ufw prints, closes the pipe, ufw takes SIGPIPE,
+# and the pipeline yields 141. This file is 'set -u' with no pipefail, so
+# that 141 is discarded, grep's 0 is the pipeline's result, and the branch
+# held — the defect was latent here, never live. It was also one line from
+# live: adding 'set -o pipefail' for unrelated robustness would silently turn
+# this into #102, skipping every UFW removal on a host the operator was told
+# is wiped, with no error and no red X to see. Captured and matched with
+# [[ ]], it is correct under any future 'set' line.
+#
+# The numbered loop had the same defect for a different reason: its condition
+# was also an early-exit reader, so it could end while rules remained, and it
+# re-read un-captured to get the number. It now reads one capture per
+# iteration and breaks on absence — the re-scan is still per-delete (numbers
+# shift after each removal), just no longer racing.
+ufw_status=""
+if command -v ufw >/dev/null; then
+  # '|| true': ufw exits non-zero when it cannot read its config, and "no
+  # usable ufw here" is nothing-to-clean, not a failure to report.
+  ufw_status="$(sudo ufw status 2>/dev/null || true)"
+fi
+
+if [[ "$ufw_status" == *"Status: active"* ]]; then
   for net in boxnet claudenet; do
-    while sudo ufw status numbered | grep -q "on $net"; do
-      n="$(sudo ufw status numbered | grep -m1 "on $net" | sed -E 's/^\[ *([0-9]+)\].*/\1/')"
+    while :; do
+      numbered="$(sudo ufw status numbered 2>/dev/null || true)"
+      line="$(printf '%s\n' "$numbered" | grep -m1 "on $net" || true)"
+      [ -n "$line" ] || break
+      n="$(printf '%s\n' "$line" | sed -E 's/^\[ *([0-9]+)\].*/\1/')"
+      [ -n "$n" ] || break
       sudo ufw --force delete "$n" >/dev/null && say "deleted UFW rule on $net"
     done
   done

test/cli.sh

@@ -2006,26 +2006,42 @@ check "teardown-host: points at box uninstall when done" 0 "" \
 check "teardown-host: refuses without a TTY and names the override (#113)" 2 \
   "--yes (or BOX_YES=1) means yes" env -u BOX_YES bash "$ROOT/host/teardown-host.sh" </dev/null
 
-# #102's race, in the one other file that sets pipefail. A daemon-free run
-# cannot exercise a UFW teardown, so the shape is pinned instead: no `ufw
-# status` may be piped into an early-exit reader here, because under this
-# file's pipefail the reader's match closes the pipe, ufw takes SIGPIPE, and
-# the branch silently reads false — skipping crumb removal on a host the
-# operator was told is clean. Both directions: the racing shape absent, the
-# capture present.
-# Comment lines are stripped before matching: the fix's own commentary quotes
+# #102's race, pinned as a CLASS rather than at the one site that had it
+# (#107). A daemon-free run cannot exercise a UFW teardown, so the shape is
+# pinned instead: nowhere under host/ or drill/ may `ufw status` be piped into
+# a reader that exits on its first match. `Status: active` is ufw's FIRST line,
+# so the reader matches, closes the pipe, ufw takes SIGPIPE, and the pipeline
+# yields 141 — under pipefail the branch silently reads false and the whole
+# firewall block is skipped on a host the operator was told is clean.
+#
+# Swept, not per-file, because absence of pipefail is what made drill/wipe.sh
+# survive the same shape: a file is only ever one `set -o pipefail` — the kind
+# of robustness tweak that sails through review — from being #102 again. The
+# sweep closes the class, so a new host/ or drill/ script inherits the pin for
+# free instead of being one more site someone has to remember.
+# Comment lines are stripped before matching: each fix's own commentary quotes
 # the racing shape to explain it, and a pin that cannot tell prose from code
 # would fail on the very comment documenting why it exists.
 # shellcheck disable=SC2016  # "$1" is the subshell's positional, passed below
-check "teardown-host: no 'ufw status' piped into an early-exit reader" 0 "" \
-  bash -c 'grep -vE "^[[:space:]]*#" "$1" | grep -qE "ufw status[^|]*\| *grep" && exit 1; exit 0' \
-    _ "$ROOT/host/teardown-host.sh"
-# shellcheck disable=SC2016  # the $-strings are literals in the target file
-check "teardown-host: the UFW branch reads a captured snapshot" 0 "" \
-  grep -qF 'if [[ "$ufw_status" == *"Status: active"* ]]; then' "$ROOT/host/teardown-host.sh"
-# shellcheck disable=SC2016  # ditto
-check "teardown-host: the numbered-delete loop breaks on absence, not on a pipe" 0 "" \
-  grep -qF '[ -n "$line" ] || break' "$ROOT/host/teardown-host.sh"
+check "no 'ufw status' is piped into an early-exit reader under host/ or drill/" 0 "" \
+  bash -c 'bad=""
+    for f in "$1"/host/*.sh "$1"/drill/*.sh; do
+      grep -vE "^[[:space:]]*#" "$f" | grep -qE "ufw status[^|]*\| *grep" && bad="$bad ${f#"$1"/}"
+    done
+    [ -z "$bad" ] || { printf "racing ufw reads in:%s\n" "$bad"; exit 1; }' \
+    _ "$ROOT"
+
+# The other direction, per file that removes UFW rules: the capture present and
+# the delete loop breaking on absence, so the sweep above cannot be satisfied by
+# deleting the block instead of fixing it.
+for f in host/teardown-host.sh drill/wipe.sh; do
+  # shellcheck disable=SC2016  # the $-strings are literals in the target files
+  check "$f: the UFW branch reads a captured snapshot" 0 "" \
+    grep -qF 'if [[ "$ufw_status" == *"Status: active"* ]]; then' "$ROOT/$f"
+  # shellcheck disable=SC2016  # ditto
+  check "$f: the numbered-delete loop breaks on absence, not on a pipe" 0 "" \
+    grep -qF '[ -n "$line" ] || break' "$ROOT/$f"
+done
 check "drill: reads the installed tree through current/" 0 "" \
   grep -qF '.local/share/box/current/VERSION' "$ROOT/drill/drill.sh"