fix(bin): carry no-mistakes ships into validation

AGENTS.md

@@ -357,7 +357,7 @@ After an autonomous merge, give the captain a one-line full-URL or local-main ou
 
 ### Validate
 
-For a no-mistakes ship, trigger validation on the same worker after its implementation commit, using the harness invocation owned by `harness-adapters`.
+For a no-mistakes ship, the brief has the same worker carry straight from its implementation commit into validation without waiting, so expect no handoff pause and no mid-task `done:` line; trigger validation yourself with the harness invocation owned by `harness-adapters` only when a worker has stopped without starting the run.
 The task worker that starts a no-mistakes run drives the pipeline and owns every `no-mistakes axi run` and `no-mistakes axi respond` call through the next gate or outcome.
 Firstmate never invokes `no-mistakes axi respond` for a crew-owned run.
 When the captain adds or changes an ask mid-task, append the captain's words to that brief's `## Captain's intent` and steer the worker; Firstmate build constraints stay in `## Firstmate spec` or the steer.

bin/fm-dod-lib.sh

@@ -218,9 +218,9 @@ EOF
       cat <<EOF
 # Definition of done
 Delivery contract: mode=no-mistakes
-The task is complete only when committed on your branch.
-When you believe it is complete, append \`done: {summary}\` to the status file and stop.
-Firstmate will then instruct you to run /no-mistakes to validate and ship a PR.
+This task ships **no-mistakes**: you validate and raise the PR yourself, in one continuous pass.
+When the implementation is committed on your branch, append \`working: implemented, starting validation\` to the status file and invoke /no-mistakes right away - do not stop and do not wait to be told to start it.
+The terminal report below is the only \`done:\` line this task ever gets.
 
 You drive no-mistakes by responding to its gates, not by implementing fixes.
 Follow the guidance no-mistakes itself provides for the mechanics: it loads when you invoke /no-mistakes, and \`no-mistakes axi run --help\` plus the \`help\` lines in each \`axi\` response are authoritative and version-matched to the installed binary.

tests/fm-brief.test.sh

@@ -260,7 +260,7 @@ test_ship_mode_is_explicit_not_registry() {
   brief="$home/data/brief-explicit-a5/brief.md"
   grep -qx "Delivery contract: mode=no-mistakes" "$brief" \
     || fail "registered direct-PR posture overrode the explicit --mode"
-  assert_grep "Firstmate will then instruct you to run /no-mistakes" "$brief" \
+  assert_grep "invoke /no-mistakes right away" "$brief" \
     "explicit no-mistakes brief did not render the pipeline definition of done"
 
   # An unregistered project is not a blocker either, because nothing is looked up.
@@ -424,6 +424,57 @@ test_ask_user_escalation_format() {
   pass "fm-brief.sh: no-mistakes ask-user findings use one event plus a verbatim snapshot"
 }
 
+# A no-mistakes ship brief carries the worker straight from its implementation
+# commit into validation. Each ship mode is covered so its definition of done
+# continues to instruct exactly one terminal `done:` line.
+test_ship_briefs_never_instruct_a_mid_task_stop() {
+  local home id brief dod count
+  home="$TMP_ROOT/mid-task-stop-home"
+  mkdir -p "$home/data"
+
+  id="brief-carry-through-c1"
+  FM_HOME="$home" "$ROOT/bin/fm-brief.sh" "$id" some-proj --mode no-mistakes >/dev/null 2>&1
+  brief="$home/data/$id/brief.md"
+  assert_present "$brief" "no-mistakes brief was not scaffolded"
+  assert_no_grep "When you believe it is complete, append \`done: {summary}\` to the status file and stop." "$brief" \
+    "no-mistakes brief still instructs a mid-task stop after the implementation commit"
+  assert_no_grep "Firstmate will then instruct you to run /no-mistakes" "$brief" \
+    "no-mistakes brief still waits for a firstmate handoff before validating"
+  assert_grep "invoke /no-mistakes right away - do not stop and do not wait to be told to start it" "$brief" \
+    "no-mistakes brief lost the carry-through instruction into validation"
+  assert_grep "append \`working: implemented, starting validation\`" "$brief" \
+    "no-mistakes brief must report the implementation milestone with a nonterminal state"
+  # The surviving `done:` must be the terminal CI-green report, and the
+  # hard-won gate rules around it must be untouched.
+  assert_grep "append \`done: PR {url} checks green\` and stop. You are finished." "$brief" \
+    "no-mistakes brief lost its terminal CI-green report"
+  assert_grep "ask-user findings are never yours to answer" "$brief" \
+    "no-mistakes brief lost the ask-user escalation rule"
+  assert_grep "NEVER pass \`--yes\` (or \`-y\`)" "$brief" \
+    "no-mistakes brief lost the --yes prohibition"
+  assert_grep "Do not hand-edit, commit, or fix findings yourself while a run is active" "$brief" \
+    "no-mistakes brief lost the active-run hands-off rule"
+  assert_grep "do not wait for it to keep monitoring in the background until merge" "$brief" \
+    "no-mistakes brief lost its CI-green return point"
+
+  # Each mode's definition of done instructs exactly one `done:` line, so no
+  # ship brief can ask for a terminal state twice for one task.
+  local mode ids
+  ids=0
+  for mode in no-mistakes direct-PR local-only; do
+    ids=$((ids + 1))
+    id="brief-single-done-c$ids"
+    FM_HOME="$home" "$ROOT/bin/fm-brief.sh" "$id" some-proj --mode "$mode" >/dev/null 2>&1
+    brief="$home/data/$id/brief.md"
+    assert_present "$brief" "$mode brief was not scaffolded"
+    dod=$(sed -n '/^# Definition of done$/,$p' "$brief")
+    count=$(printf '%s\n' "$dod" | grep -c -F 'append `done:' || true)
+    [ "$count" -eq 1 ] \
+      || fail "$mode definition of done instructs $count \`done:\` lines, expected exactly 1"
+  done
+  pass "fm-brief.sh: ship briefs carry through to their single terminal report"
+}
+
 test_ship_project_memory_wording() {
   local home id brief
   home="$TMP_ROOT/project-memory-home"
@@ -912,6 +963,7 @@ test_delivery_flags_are_refused_where_they_do_not_apply
 test_faster_paths_use_configured_authority_without_stacked_review
 test_no_mistakes_dod_wording
 test_ask_user_escalation_format
+test_ship_briefs_never_instruct_a_mid_task_stop
 test_ship_project_memory_wording
 test_herdr_lab_contract_is_explicit_and_complete
 test_herdr_lab_contract_quotes_foreign_firstmate_path

tests/fm-watch-triage.test.sh

@@ -174,7 +174,22 @@ record_pi_busy() {  # <state-dir> <id>
     --source pi-ext --event agent-start
 }
 
-reap() { kill "$1" 2>/dev/null || true; wait "$1" 2>/dev/null || true; }
+# Teardown only: callers have finished all assertions about this watcher.
+# fm-watch.sh can defer TERM while foreground work unwinds, so bound its grace
+# period and then escalate to KILL to keep test teardown deterministic.
+reap() {
+  local pid=$1 i=0
+  kill "$pid" 2>/dev/null || true
+  # 100 ticks matches the wait_for_exit budget documented above: far longer than
+  # a healthy watcher needs to honour TERM, short enough to stay bounded.
+  while [ "$i" -lt 100 ]; do
+    kill -0 "$pid" 2>/dev/null || break
+    sleep 0.1
+    i=$((i + 1))
+  done
+  kill -9 "$pid" 2>/dev/null || true
+  wait "$pid" 2>/dev/null || true
+}
 
 # --- pure classifier predicates (fm-classify-lib.sh) ------------------------