fix(tmux): confirm delivery by draft transition, not by prompt glyph (#1257)

send-message.sh located the REPL input box with grep -E '❯|^>|│ >'. That set is
Claude Code's box. A pi seat renders a bare U+2500 rule with no glyph, so on every
idle pi seat the capture succeeded, the grep matched nothing, status stayed
"unconfirmed", and the tool exited 2 "may be UNDELIVERED" with the paste and the
Enter both landed. The stderr tells the operator to retry, and that retry is the
duplicate delivery reported against the same tool.

Confirmation is now runtime-agnostic: our message tail sits on the input line
(located by cursor row, no glyph) before Enter and has left it after. That
transition is positive proof of submission.

Absence still proves nothing, which is the guard the 2026-08 fix was reaching for
and got backwards. Two positive checks keep it:

  - a prompt box that IS locatable and still carries our tail => draft, exit 2.
    This covers the cursor-row blind spot: a cooked pane whose foreground process
    never reads stdin echoes the paste through the kernel line discipline and
    moves the cursor off it on Enter, which by cursor row alone is indistinguishable
    from a real submit.
  - no draft ever observed on the input line => unconfirmed, non-zero.

Tests, both red-first against the shipping blob d397907:

  test-send-message-glyph-agnostic.sh (new, 6 fixtures)  4/6 -> 6/6
  test-send-message-verdict.sh (fixture 2 reshaped, 2b added)  3/4 -> 4/4

Fixture 2 of the verdict suite asserted exit 2 for a glyphless pane that submits
and was labelled "false-positive FIXED". A pi seat is that fixture, so the suite
was locking the bug in. It is reshaped deliberately, and the guard it was credited
with moves to new fixture 2b (glyphless AND non-submitting, raw/no-echo) so the
"never infer delivered from absence" property is tested positively rather than as
a side effect.

Measured on tmux 3.7b (sb-it-1-dt), 3.5a (fomo-lin), and dragon-lin.

Co-authored-by: scooby <[email protected]>

packages/mosaic/framework/tools/quality/test-enumeration-exclusions.txt

@@ -39,20 +39,3 @@ packages/mosaic/framework/tools/tmux/test-send-message-verdict.sh | requires rea
 # recorded judgement. These lines ARE that judgement, signed.)
 packages/mosaic/framework/tools/orchestrator/smoke-test.sh | behavior smoke checks for coord continue/run workflows, run manually by orchestrator seats; unmeasured in CI; #1017 burndown
 packages/mosaic/framework/tools/wake/validate-973/microtest-wake-assert.sh | #973 instrument self-test, run as a precondition of the validate-973 evidence procedure rather than as a standing CI suite; #1017 burndown candidate
-
-# --- tools/fleet: precondition is unsatisfiable in the CI image (#1271) ---
-# Signed by fred (sb-it-1-dt, 2026-08-16) at origin/next 476db12.
-# This suite asserts the launcher's behaviour when `mosaic` and `pi` are MISSING.
-# It shims fakes into $FAKE_BIN, but the constructed PANE_PATH always ends in the
-# real system path, so on a host that installs those binaries the missing-binary
-# cases cannot be measured at all. The suite's own guard (line 103) says so and
-# fails rather than reporting a pass it cannot back. That guard is correct.
-# The error was wiring the suite into CI: #1017 (c56483eb) enumerated it and
-# dropped this exclusion, and the CI image provides `pi` in the system path, so
-# it has failed on every pipeline since. Measured 2026-08-16 across pipelines
-# 2444 (#1256), 2438 (#1240) and 2441 (#1017-quality): exactly one FAIL line in
-# each full log, identical, this assertion; control `zzz-not-present-zzz` -> 0.
-# Burn-down and the full measurement are tracked in #1271; unwired by PR #1270.
-# Because test:framework-shell is one && chain and this sat at position 44 of 48,
-# the four suites after it had not run at all since the merge.
-packages/mosaic/framework/tools/fleet/test-start-agent-session.sh | precondition unsatisfiable in the CI image: asserts missing-binary behaviour, but PANE_PATH always ends in the system path and the image provides `pi` there; guard at line 103 fails by design rather than passing unmeasured. Burn down by controlling the tail of PANE_PATH inside the test. NOT by removing `pi` from the image: the CI image installs @earendil-works/[email protected] deliberately (measured in pipeline 2444's test-step log), and other suites depend on that pin. Burn-down tracked in #1271

packages/mosaic/framework/tools/tmux/send-message.sh

@@ -97,13 +97,34 @@ printf '%s' "$MSG" | "${tmux_cmd[@]}" load-buffer -b "$BUF" -
 #   would otherwise accumulate forever.
 sleep 0.5
 
-# 2) Submit, then POSITIVELY confirm submission; flush with another Enter if it is
+# 2) Submit, then POSITIVELY confirm submission by DRAFT TRANSITION, not by prompt
-#    still a draft. Success requires positive evidence — the queued banner, OR the
+#    glyph. The historical bug was treating ABSENCE of a draft as delivery; the
-#    REPL input box located AND clear of our message tail. The historical bug was
+#    2026-08 fix over-corrected to glyph inference (grep '❯|^>|│ >'), which locates
-#    treating ABSENCE of a draft as delivery: if the prompt glyph was never matched
+#    only Claude Code's box and false-NEGATIVES every glyphless REPL (pi renders a
-#    (wrong pane / prompt-glyph drift), an unsubmitted message read as "delivered"
+#    U+2500 rule, no glyph) — a delivered message reported "UNDELIVERED", driving a
-#    and worker->lead relays stalled silently. We now default to UNCONFIRMED and only
+#    retry that duplicates it. Runtime-agnostic evidence: our message tail sits on
-#    upgrade to delivered on positive evidence; anything we cannot confirm fails loud.
+#    the INPUT line (located by the cursor row, not a glyph) BEFORE Enter, and has
+#    LEFT it AFTER — that transition is positive proof of submission and needs no
+#    glyph. Absence alone still never means delivered: if we never saw our draft on
+#    the input line we stay UNCONFIRMED (wrong/dead pane), and a draft that never
+#    leaves the input line stays a DRAFT (exit 2), preserving both historical guards.
+_cursor_line() { # echo the pane's current input (cursor) line, glyph-free
+  local cy line
+  cy=$("${tmux_cmd[@]}" display-message -p -t "$EFFECTIVE_TARGET" -F '#{cursor_y}' 2>/dev/null) || return 1
+  [ -n "$cy" ] || return 1
+  "${tmux_cmd[@]}" capture-pane -t "$EFFECTIVE_TARGET" -p 2>/dev/null | sed -n "$((cy + 1))p"
+}
+_draft_on_input() { # true iff our message tail is sitting on the input line now
+  [ -n "$snippet" ] || return 1
+  printf '%s' "$(_cursor_line)" | grep -qF "$snippet"
+}
+
+# Baseline: after the paste, our draft must be on the input line. This is positive
+# proof we are on the right pane and the paste landed — the anchor the transition
+# check measures against.
+saw_draft=0
+_draft_on_input && saw_draft=1
+
 status="unconfirmed"
 for attempt in $(seq 1 $((RETRIES + 1))); do
   "${tmux_cmd[@]}" send-keys -t "$EFFECTIVE_TARGET" Enter
@@ -113,20 +134,26 @@ for attempt in $(seq 1 $((RETRIES + 1))); do
   if printf '%s' "$pane" | grep -qF "$QUEUED_RE"; then
     status="queued"; break
   fi
-  # Locate the REPL input box (prompt glyph). If we cannot see it, we have NO
+  # POSITIVE draft evidence from a located prompt box, when one exists. This is the
-  # evidence of submission state — stay UNCONFIRMED and retry; never infer delivery.
+  # cursor-row check's blind spot: a pane in COOKED mode (a plain shell whose
+  # foreground process never reads stdin) echoes our paste via the kernel line
+  # discipline and moves the cursor off it on Enter, which is indistinguishable from
+  # a real submit by cursor row alone. If a prompt box IS locatable and still carries
+  # our tail, that is affirmative proof the message was not consumed. Absence of a
+  # glyph is still never used for anything — that inference is the original E7 bug.
   promptline=$(printf '%s' "$pane" | grep -E '❯|^>|│ >' | tail -1)
-  if [ -z "$promptline" ]; then
+  if [ -n "$promptline" ] && [ -n "$snippet" ] && printf '%s' "$promptline" | grep -qF "$snippet"; then
-    status="unconfirmed"; continue
-  fi
-  # Input box located AND still carrying our tail => unsubmitted draft. Flush + retry.
-  # (Submitted messages scroll up into history; a draft stays on the ❯ line.)
-  if [ -n "$snippet" ] && printf '%s' "$promptline" | grep -qF "$snippet"; then
     status="draft"; continue
   fi
-  # Input box located AND clear of our tail => positively submitted. This is the
+  if [ "$saw_draft" = 1 ]; then
-  # only path to success besides the queued banner.
+    if _draft_on_input; then
-  status="delivered"; break
+      status="draft"; continue      # still on the input line => not submitted; flush + retry
+    fi
+    status="delivered"; break        # left the input line => positively submitted
+  fi
+  # No confirmed baseline yet: try to (re)acquire it; never infer delivery from absence.
+  if _draft_on_input; then saw_draft=1; status="draft"; continue; fi
+  status="unconfirmed"; continue
 done
 
 [ "$VERBOSE" = 1 ] && { echo "--- pane tail ($TARGET) ---"; printf '%s\n' "$pane" | tail -4; echo "---"; }

packages/mosaic/framework/tools/tmux/test-send-message-glyph-agnostic.sh

@@ -0,0 +1,97 @@
+#!/usr/bin/env bash
+# Red-first regression test for E7 (#1017 task 2): the confirm-check must bind
+# "delivered" to WHETHER THE MESSAGE WAS SUBMITTED, not to which runtime's prompt
+# glyph is present. A pi seat renders a U+2500 rule input box with no ❯/^>/│ >
+# glyph; send-message.sh:118 locates the box only by glyph, so a genuinely
+# delivered message on a glyphless REPL falsely reports exit 2 "may be UNDELIVERED",
+# and the operator's rc=2-driven retry duplicates it.
+#
+# Parameterized on $SEND: RED against the shipping blob (B and D fail), GREEN
+# against a candidate patch. No pi; no fake HOME; hermetic throwaway socket.
+#
+# Submission counting is EXACT and terminal-echo-independent: the fixture message
+# is `echo <tok> >>SINK`; each real submission appends one line. wc -l SINK ==
+# number of times the REPL actually executed the send. This does not depend on how
+# many times the marker string is painted on screen.
+set -u
+SEND="${SEND:?set SEND=/path/to/send-message.sh}"
+SOCKET="glyphagnostic-$$"
+TMP="$(mktemp -d)"
+tmux() { command tmux -L "$SOCKET" "$@"; }
+cleanup() { command tmux -L "$SOCKET" kill-server 2>/dev/null; rm -rf "$TMP"; }
+trap cleanup EXIT
+pass=0; fail=0
+ok() { printf 'ok   %s\n' "$1"; pass=$((pass+1)); }
+no() { printf 'FAIL %s -- %s\n' "$1" "$2"; fail=$((fail+1)); }
+
+mk() { tmux new-session -d -s "$1" -x 120 -y 40 -c "$TMP" "PS1='$2' exec bash --noprofile --norc -i"; sleep 0.5; }
+subs() { [ -f "$1" ] && wc -l <"$1" | tr -d ' ' || echo 0; }  # exact submission count
+
+echo "SEND=$SEND    tmux $(command tmux -V | awk '{print $2}')"
+
+# --- A (control): glyph box (❯) that submits => exit 0, exactly one submission.
+mk ctl '❯ '
+SINK="$TMP/sink.ctl"
+out=$("$SEND" -L "$SOCKET" -t ctl -m "echo x >>'$SINK'" 2>"$TMP/e.ctl"); rc=$?; sleep 0.4
+if [ "$rc" = 0 ] && [ "$(subs "$SINK")" = 1 ]; then
+  ok "control: ❯-box submits => exit 0, exactly one submission"
+else no "control: ❯-box submits => exit 0, one submission" "rc=$rc subs=$(subs "$SINK") err=[$(cat "$TMP/e.ctl")]"; fi
+
+# --- B (THE false-rc regression): glyphless U+2500 box that SUBMITS. Message lands
+#     (subs==1) yet shipping reports exit 2. Must be exit 0.
+mk sub $'──────── \n'
+SINK="$TMP/sink.sub"
+out=$("$SEND" -L "$SOCKET" -t sub -m "echo x >>'$SINK'" 2>"$TMP/e.sub"); rc=$?; sleep 0.4
+if [ "$rc" = 0 ] && [ "$(subs "$SINK")" = 1 ]; then
+  ok "glyphless: U+2500 box that submits => exit 0 (delivered, not 'UNDELIVERED')"
+else no "glyphless: U+2500 box that submits => exit 0" \
+        "rc=$rc subs=$(subs "$SINK")(delivered=$([ "$(subs "$SINK")" -ge 1 ] && echo yes||echo no)) err=[$(cat "$TMP/e.sub")]"; fi
+
+# --- D (duplicate arm): operator follows the rc=2 stderr and retries once. On the
+#     glyphless box, shipping => two submissions (the reported duplicate). The
+#     property: one logical send => exactly one submission. Same fix closes it.
+mk dup $'──────── \n'
+SINK="$TMP/sink.dup"
+tries=0
+for attempt in 1 2; do
+  tries=$((tries+1))
+  out=$("$SEND" -L "$SOCKET" -t dup -m "echo x >>'$SINK'" 2>/dev/null); rc=$?
+  sleep 0.4
+  [ "$rc" = 0 ] && break   # operator stops retrying only when told delivered
+done
+if [ "$(subs "$SINK")" = 1 ]; then
+  ok "duplicate: one logical send (rc-driven retry) => exactly one submission (tries=$tries)"
+else no "duplicate: one logical send => exactly one submission" "submissions=$(subs "$SINK") tries=$tries"; fi
+
+# --- E (faithful hung managed TUI, NOT a cooked shell): raw/no-echo, paints nothing.
+#     A cooked `sleep infinity` echoes the paste via the kernel line discipline and
+#     false-passes a cursor-row fix that is correct on real seats (measured). So: raw.
+mk_rawstuck() { tmux new-session -d -s "$1" -x 120 -y 40 -c "$TMP" \
+  "bash --noprofile --norc -c 'stty -echo -icanon min 1 time 0 2>/dev/null; exec sleep infinity'"; sleep 0.5; }
+mk_rawstuck estuck
+SINK="$TMP/sink.estuck"
+out=$("$SEND" -L "$SOCKET" -t estuck -r 1 -m "this stuck draft was never submitted" 2>/dev/null); rc=$?
+sleep 0.3
+if [ "$rc" != 0 ] && [ "$(subs "$SINK")" = 0 ]; then
+  ok "raw/no-echo stuck TUI (not submitted) => non-zero (no false delivered)"
+else no "raw stuck TUI must NOT report delivered" "rc=$rc subs=$(subs "$SINK")"; fi
+
+# --- F (busy/queued branch, your BUSY-not-runtime finding): glyphless pane rendering the
+#     queued banner, never consuming. QUEUED_RE :113 fires before the glyph grep => rc=0.
+mk_busy() { tmux new-session -d -s "$1" -x 120 -y 40 -c "$TMP" \
+  "bash --noprofile --norc -c 'printf \"Press up to edit queued messages\n\"; exec sleep infinity'"; sleep 0.5; }
+mk_busy ebusy
+SINK="$TMP/sink.ebusy"
+out=$("$SEND" -L "$SOCKET" -t ebusy -m "echo x >>'$SINK'" 2>/dev/null); rc=$?; sleep 0.3
+if [ "$rc" = 0 ]; then
+  ok "busy/queued-banner glyphless => exit 0 (queued is delivery; runtime owns custody)"
+else no "busy/queued-banner must report delivered" "rc=$rc"; fi
+
+# --- C (historical-bug guard): unresolvable target. No pane ever carried our draft
+#     => must fail, never infer delivered from absence of a glyph/snippet.
+if out=$("$SEND" -L "$SOCKET" -t "nonexistent-$$" -m "echo x >>'$TMP/sink.wrong'" 2>/dev/null); then
+  no "wrong-pane: unresolvable target must NOT report success" "expected non-zero, got 0"
+else ok "wrong-pane: unresolvable target => non-zero (no false delivered)"; fi
+
+echo "---"; echo "pass=$pass fail=$fail"
+[ "$fail" = 0 ]

packages/mosaic/framework/tools/tmux/test-send-message-verdict.sh

@@ -4,10 +4,13 @@
 #
 #   1. DELIVERED   — a REPL that renders a `❯ ` input box and submits on Enter
 #                    (text scrolls to history, box clears) => exit 0 "✓ delivered".
-#   2. UNCONFIRMED — a pane with NO locatable prompt glyph. This is the exact
+#   2. DELIVERED   — a pane with NO prompt glyph that DOES submit => exit 0. A pi
-#                    historical FALSE POSITIVE: pre-patch it printed "✓ delivered"
+#                    seat is this fixture (U+2500 rule, no glyph). Reshaped for
-#                    exit 0; post-patch it MUST fail loud (exit 2, stderr
+#                    #1257; see the note at the fixture for why the old exit-2
-#                    "could not confirm submission").
+#                    assertion was wrong.
+#   2b. UNCONFIRMED— a glyphless pane that never submits (raw/no-echo hung TUI)
+#                    => must fail loud. This carries the historical
+#                    false-positive guard that fixture 2 used to be credited with.
 #   3. DRAFT       — a `❯ `-prompt pane that never submits (message stays on the
 #                    input line) => exit 2, stderr "unsubmitted draft".
 set -uo pipefail
@@ -37,19 +40,44 @@ else
   no "delivered: ❯-prompt REPL that submits => exit 0 ✓ delivered" "rc=$rc out=[$out] err=[$(cat "$TMP/e1")]"
 fi
 
-# --- Fixture 2: NO prompt glyph (default bash PS1). THE regression: pre-patch this
+# --- Fixture 2: NO prompt glyph, and the pane DOES submit (interactive bash).
-#     was a silent false-positive "delivered"; post-patch it must be unconfirmed→exit 2.
+#     RESHAPED 2026-08-16 (#1257), deliberately. This fixture previously asserted
+#     exit 2 here and was labelled "false-positive FIXED". That assertion was wrong,
+#     and locking it in is what kept E7 alive: the pane submits, so "delivered" is
+#     the truth, and a pi seat — whose input box is a bare U+2500 rule with no glyph
+#     — IS this fixture. Reporting exit 2 for it told operators a delivered message
+#     may be undelivered, and the retry that advice invites is the duplicate.
+#
+#     The guard this fixture was reaching for is real and is NOT dropped: "never
+#     infer delivered from absence" is now enforced positively by fixture 2b below
+#     (glyphless AND not submitting => must fail) and by fixture 3 (locatable box
+#     still carrying our tail => draft). Absence alone decides nothing either way.
 tmux -L "$SOCKET" new-session -d -s noglyph -c "$TMP" \
   'PS1="sh-noglyph$ " exec bash --noprofile --norc -i'
 sleep 0.3
-if out=$("$SEND" -L "$SOCKET" -t "=noglyph" -m "verdict fixture two must fail loud" 2>"$TMP/e2"); then
+out=$("$SEND" -L "$SOCKET" -t "=noglyph" -m "verdict fixture two must fail loud" 2>"$TMP/e2"); rc=$?
-  no "unconfirmed: glyphless pane must NOT report success" "expected exit 2, got 0 (out=[$out])"
+if [ "$rc" -eq 0 ] && printf '%s' "$out" | grep -qF "✓ delivered"; then
+  ok "delivered: glyphless pane that submits => exit 0 (runtime-agnostic, E7 FIXED)"
+else
+  no "delivered: glyphless pane that submits => exit 0" "rc=$rc out=[$out] err=[$(cat "$TMP/e2")]"
+fi
+
+# --- Fixture 2b: NO prompt glyph AND never submits — a hung managed TUI holding the
+#     terminal in raw/no-echo, which is what a stuck agent seat actually is (measured
+#     on live pi: stty -echo -icanon). Nothing is echoed, nothing is consumed, so
+#     there is no positive evidence of submission and the tool MUST fail loud. This
+#     is the historical false-positive guard, kept as a positive test.
+tmux -L "$SOCKET" new-session -d -s rawstuck -c "$TMP" \
+  'bash --noprofile --norc -c "stty -echo -icanon min 1 time 0 2>/dev/null; exec sleep infinity"'
+sleep 0.3
+if out=$("$SEND" -L "$SOCKET" -t "=rawstuck" -r 1 -m "verdict fixture two-b never submitted" 2>"$TMP/e2b"); then
+  no "unconfirmed: glyphless hung TUI must NOT report success" "expected non-zero, got 0 (out=[$out])"
 else
   rc=$?
-  if [ "$rc" -eq 2 ] && grep -qF "could not confirm submission" "$TMP/e2"; then
+  if [ "$rc" -ne 0 ] && grep -qF "could not confirm submission" "$TMP/e2b"; then
-    ok "unconfirmed: glyphless pane => exit 2 + 'could not confirm submission' (false-positive FIXED)"
+    ok "unconfirmed: glyphless hung TUI (raw/no-echo) => non-zero + 'could not confirm submission'"
   else
-    no "unconfirmed: glyphless pane => exit 2 + stderr" "rc=$rc err=[$(cat "$TMP/e2")]"
+    no "unconfirmed: glyphless hung TUI => non-zero + stderr" "rc=$rc err=[$(cat "$TMP/e2b")]"
   fi
 fi
 

packages/mosaic/package.json

@@ -25,7 +25,7 @@
     "lint": "eslint src",
     "typecheck": "tsc --noEmit",
     "test": "vitest run --passWithNoTests && pnpm run test:framework-shell",
-    "test:framework-shell": "bash framework/tools/quality/scripts/check-test-enumeration.sh && bash framework/tools/quality/scripts/test-check-test-enumeration.sh && python3 src/lease-broker/daemon_deadline_unittest.py && python3 src/lease-broker/normative_fragments_unittest.py && python3 src/lease-broker/promotion_binding_unittest.py && python3 src/lease-broker/promotion_trigger_unittest.py && python3 src/lease-broker/receipt_challenge_unittest.py && python3 src/lease-broker/context_recovery_unittest.py && python3 src/lease-broker/recovery_runtime_unittest.py && python3 src/lease-broker/recovery_b1_adversarial_unittest.py && python3 src/lease-broker/receipt_observer_client_unittest.py && python3 src/lease-broker/invariant_r_unittest.py && python3 src/lease-broker/framework_skill_portability_unittest.py && python3 src/mutator-gate/runtime_tools_unittest.py && python3 src/mutator-gate/runtime_launch_guard_unittest.py && python3 src/mutator-gate/version_coupling_unittest.py && python3 framework/tools/lease-broker/check-runtime-launches.py --root ../.. && bash framework/tools/codex/test-pr-diff-context.sh && bash framework/tools/qa/test-deps-preflight.sh && bash framework/tools/git/test-pr-review-gitea-comment.sh && bash framework/tools/git/test-pr-review-repo-host-override.sh && bash framework/tools/git/test-ci-queue-wait-branch-absent.sh && bash framework/tools/git/test-ci-queue-wait-tristate.sh && bash framework/tools/git/test-ci-queue-wait-github-checks.sh && bash framework/tools/git/test-pr-merge-queue-branch.sh && bash framework/tools/git/test-pr-merge-head-pin.sh && bash framework/tools/git/test-pr-merge-message-field.sh && bash framework/tools/git/test-git-credential-mosaic.sh && bash framework/tools/git/test-gitea-token-identity.sh && bash framework/tools/woodpecker/test-terminal-green-contract.sh && bash framework/tools/_scripts/test-install-ordering-guard.sh && bash framework/tools/_scripts/test-mosaic-init-rce.sh && bash framework/tools/tmux/agent-send.test.sh && bash framework/tools/wake/test-wake-store-ack.sh && bash framework/tools/wake/test-wake-store-enqueue-race.sh && bash framework/tools/wake/test-wake-digest-hmac.sh && bash framework/tools/wake/test-wake-digest-quarantine.sh && bash framework/tools/wake/test-wake-detector.sh && bash framework/tools/wake/test-wake-fn-oracle.sh && bash framework/tools/wake/test-wake-reconcile.sh && bash framework/tools/wake/test-wake-beacon.sh && bash framework/tools/wake/test-wake-preimage.sh && bash framework/tools/wake/test-wake-install.sh && bash framework/tools/glpi/test-list-http-status.sh && bash framework/tools/orchestrator/test-board-roll.sh && bash framework/tools/woodpecker/test-ci-wait-exit-matrix.sh && bash framework/tools/_scripts/test-fleet-transport-check.sh"
+    "test:framework-shell": "bash framework/tools/quality/scripts/check-test-enumeration.sh && bash framework/tools/quality/scripts/test-check-test-enumeration.sh && python3 src/lease-broker/daemon_deadline_unittest.py && python3 src/lease-broker/normative_fragments_unittest.py && python3 src/lease-broker/promotion_binding_unittest.py && python3 src/lease-broker/promotion_trigger_unittest.py && python3 src/lease-broker/receipt_challenge_unittest.py && python3 src/lease-broker/context_recovery_unittest.py && python3 src/lease-broker/recovery_runtime_unittest.py && python3 src/lease-broker/recovery_b1_adversarial_unittest.py && python3 src/lease-broker/receipt_observer_client_unittest.py && python3 src/lease-broker/invariant_r_unittest.py && python3 src/lease-broker/framework_skill_portability_unittest.py && python3 src/mutator-gate/runtime_tools_unittest.py && python3 src/mutator-gate/runtime_launch_guard_unittest.py && python3 src/mutator-gate/version_coupling_unittest.py && python3 framework/tools/lease-broker/check-runtime-launches.py --root ../.. && bash framework/tools/codex/test-pr-diff-context.sh && bash framework/tools/qa/test-deps-preflight.sh && bash framework/tools/git/test-pr-review-gitea-comment.sh && bash framework/tools/git/test-pr-review-repo-host-override.sh && bash framework/tools/git/test-ci-queue-wait-branch-absent.sh && bash framework/tools/git/test-ci-queue-wait-tristate.sh && bash framework/tools/git/test-ci-queue-wait-github-checks.sh && bash framework/tools/git/test-pr-merge-queue-branch.sh && bash framework/tools/git/test-pr-merge-head-pin.sh && bash framework/tools/git/test-pr-merge-message-field.sh && bash framework/tools/git/test-git-credential-mosaic.sh && bash framework/tools/git/test-gitea-token-identity.sh && bash framework/tools/woodpecker/test-terminal-green-contract.sh && bash framework/tools/_scripts/test-install-ordering-guard.sh && bash framework/tools/_scripts/test-mosaic-init-rce.sh && bash framework/tools/tmux/agent-send.test.sh && bash framework/tools/wake/test-wake-store-ack.sh && bash framework/tools/wake/test-wake-store-enqueue-race.sh && bash framework/tools/wake/test-wake-digest-hmac.sh && bash framework/tools/wake/test-wake-digest-quarantine.sh && bash framework/tools/wake/test-wake-detector.sh && bash framework/tools/wake/test-wake-fn-oracle.sh && bash framework/tools/wake/test-wake-reconcile.sh && bash framework/tools/wake/test-wake-beacon.sh && bash framework/tools/wake/test-wake-preimage.sh && bash framework/tools/wake/test-wake-install.sh && bash framework/tools/fleet/test-start-agent-session.sh && bash framework/tools/glpi/test-list-http-status.sh && bash framework/tools/orchestrator/test-board-roll.sh && bash framework/tools/woodpecker/test-ci-wait-exit-matrix.sh && bash framework/tools/_scripts/test-fleet-transport-check.sh"
   },
   "dependencies": {
     "@mosaicstack/brain": "workspace:*",

scratchpads/2026-08-17-fleet-identity-comms-mosaic-tree.md

@@ -1,73 +0,0 @@
-# 2026-08-17 — Fleet identity, comms delivery, and the ~/.mosaic tree (continuation record)
-
-> **Status:** active continuation record | **Owner:** Jason (rulings) / fleet (delivery) | **Created:** 2026-08-17, sb-it-1-dt session with Jarvis (jarvis-brain)
-> **Audience:** the homelab agents continuing this effort tonight. Read this whole file before acting; it supersedes nothing but preserves structure and decisions that must not be lost.
-
----
-
-## Why this exists
-
-A session on sb-it-1-dt (2026-08-17) produced three architecture decisions (two awaiting Jason's ruling), one incident postmortem (#1295), interim guardrail edits in the user-owned `~/.mosaic/` contract tree, and one new tool (`ensure-watcher.sh`). The work spans jarvis-brain (P0, not retained) and this repo (the product). **This file is the stack-side anchor so continuation does not depend on jarvis-brain surviving.**
-
-## 1. The `~/.mosaic` tree model — as-built, preserve this structure
-
-Three-tree split (this is design intent, not accident; keep it through all framework work):
-
-| Tree                | Owner     | Rule                                                                                                          |
-| ------------------- | --------- | ------------------------------------------------------------------------------------------------------------- |
-| `~/.config/mosaic/` | framework | upgrade-managed templates; NEVER user-edited; `mosaic upgrade` may overwrite                                  |
-| `~/.mosaic/`        | user      | working contracts, guides, fleet agents; upgrades reconcile with **deny-wins** (user edits never overwritten) |
-| repo satellites     | repos     | bootstrapped per-repo `.mosaic/` state                                                                        |
-
-As-built inventory of `~/.mosaic` on sb-it-1-dt (2026-08-17):
-
-- **Contract core:** `CONSTITUTION.md` (L0 law), `AGENTS.md` (dispatcher + guide router + Fleet Comms Watcher requirement), `SOUL.md` (generic base for ALL fleet agents, zero persona — includes the new **Fleet Boundaries** section), `STANDARDS.md` (universal standards — includes new **session identity** + **comms watcher hygiene** sections), `SYSTEM.md` (pure communication contract, byte-identical to jarvis-brain's prompt-testing `sr_opus_5_system_prompt.md`), `USER.md`, `TOOLS.md`.
-- **`guides/`** — user-owned working copies (E2E-DELIVERY, ORCHESTRATOR(+PROTOCOL,+LEARNINGS), WAKE-DOCTRINE, VAULT-SECRETS, etc.).
-- **`fleet/agents/`** — the per-agent store (this is MOSAIC-D-002's substrate, already in use):
-  - real agent dirs: `fargo/`, `orchestrator/`, `probe/`, `vision/`, `weekly-update/` — shape: `profile.json` (harness/account/overlay pointer) + `overlay.json` + `SOUL.md` (persona) + `scratch/` `work/` `notes/` subdirs (hygiene rules in root SOUL.md)
-  - `*.env.generated` launch overlays: `luna` `sol` `terra` (carry `MOSAIC_AGENT_NAME`, `_CLASS`, `_RUNTIME`, `_MODEL`, `_REASONING`, `_TOOL_POLICY`) — these are the mosaic-fleet seat launch envs; `inbox.env`, `itops.env` also present
-  - `probe/` is the validated layout proof: auth-bundle symlink chain, per-agent sessions, plugin-store symlink (from 2026-08-07)
-- **`auth/`, `config/`, `memory/`, `plugins/`, `skills/`, `skills-local/`** — per-tree copies/links for runtime isolation.
-- Related but outside the tree: watcher units at `~/.config/systemd/user/<agent>-comms-watcher.service`; watcher seen-state at `~/.local/state/comms-watcher-<agent>/`.
-
-## 2. Decisions register (2026-08-17 session)
-
-Full strict records live in jarvis-brain `docs/decisions/mosaic-stack/` (render on its dashboard); both are **Pending Jason's ruling**. Summaries so the content survives P0:
-
-- **MOSAIC-D-001 — SYSTEM.md as canonical harness system prompt.** Static core (Constitution+AGENTS+USER+overlays) in one tracked file; launcher renders dynamic tail (mission/PRD/fleet/persona). Delivery: `--append-system-prompt` (repeatable) for pi/claude; symlinked core file for codex (`$CODEX_HOME/instructions.md`) and opencode (`AGENTS.md`); their dynamic tail via initial prompt (needs live verification). Static-first order is the cache win. `SYSTEM.md` in `~/.mosaic` today is the communication-contract file — D-001's SYSTEM.md is the broader composition; naming to reconcile at implementation.
-- **MOSAIC-D-002 — per-agent harness homes + mechanical profiles.** Launch with targeted config-dir env vars (e.g. `PI_CODING_AGENT_DIR=~/.mosaic/fleet/agents/<name>/pi`), NOT literal HOME. SOUL.md identity mechanically generated from roster (single writer; kills the hand-copy drift measured in `agents/vision/SOUL.md` on jarvis-brain: declared Jarvis, answered Vision). Composes: SYSTEM core → per-agent SOUL → dynamic tail. `MOSAIC_AGENT_NAME` stays load-bearing for comms.
-- **Comms delivery tooling belongs in the STACK framework, not jarvis-brain** (decided in discussion; supersedes the interim placement). jarvis-brain keeps only the transport _data_ (`comms/` tree) while it lives. Agents launch from their own repos (terra from `~/src/stack` etc.) — delivery is transport-repo-relative, so this works; but every installed watcher unit's ExecStart currently points into `~/src/jarvis-brain/scripts/` — that dependency is the P0 trap to remove. Migration = move tools + regenerate units, in one step.
-- **Watcher provisioning is instantiation duty, never running-agent duty.** Interim landed as jarvis-brain `scripts/comms/ensure-watcher.sh` (idempotent ensure + `--status` boot check + interim identity warnings: missing target session, pane `MOSAIC_AGENT_NAME` mismatch via `/proc/<child>/environ`, bare-runtime NOTE). Framework move: fold into `mosaic agent --new` + fleet launch + `mosaic doctor` drift check.
-- **Prose guardrails landed (interim fences until mechanical fixes):** `~/.mosaic/SOUL.md` Fleet Boundaries (wrong-session tripwire; comms ownership; cross-agent investigation requires tasking) · `STANDARDS.md` session identity + watcher hygiene · `AGENTS.md` Fleet Comms Watcher requirement (P0-interim script path marked transitional).
-
-## 3. Incident → #1295 (already tracked here)
-
-`https://git.mosaicstack.dev/mosaicstack/stack/issues/1295` — docs-seat incident: cwd-keyed session files served three lives (dev chat → goals seat → 22 watcher injections into a wedged process); name-based watcher delivery with no identity verification; wedge after pi 0.84.1→0.84.2 update passes every liveness instrument. Proposed fixes enumerated there; provisioning follow-up in comment ID 23027.
-
-## 4. Open work queue (suggested sequence)
-
-1. **Comms tooling migration PR** (lane `next`): move `comms-watcher.sh`, `install-watcher.sh`, `ensure-watcher.sh` into the framework tree → deploy `~/.config/mosaic/tools/comms/`; regenerate existing units' ExecStart to framework paths (one-command sweep); keep `COMMS_WATCH_REPO` per-host config (points at a brain checkout until the queue transport lands). Reference: jarvis-brain commit `701c353b1`.
-2. **Ensure-on-instantiation**: `mosaic agent --new` / fleet launch call ensure semantics; `mosaic doctor` gains the drift check (`--status --all` semantics + `fred`'s hand-written unit as the known drift case; also note daphne/docs/happy/pepper/sanity/tiny/fargo currently have no watcher — cover or consciously exempt).
-3. **MOSAIC-D-002 implementation** (after ruling): per-agent homes via targeted env vars; roster-generated SOUL.md single-writer; extend the existing `*.env.generated` pattern; launch ledger keeps `config_home` audit.
-4. **MOSAIC-D-001 implementation** (after ruling): SYSTEM.md sourcing + per-harness delivery + `compose-contract` becomes render-core+tail with drift check; bench cache-ordering before/after (jarvis-brain `domains/software-dev/mosaic-stack/prompt-testing` has the bench).
-5. **Queue transport + forced separation** (longer term): supersedes watcher path; identity-verification and wedge-detection remain valid regardless of transport.
-6. **Docs inheritance**: jarvis-brain AGENTS.md's durable comms guidance (E7 pi-glyph delivery gotchas, capture-pane rules, comms protocol) must be inherited into stack docs before P0 retirement.
-
-## 5. Rules for tonight's agents
-
-- Lane: **`next`** only; nothing to `main` without Jason (standing ruling).
-- Attribution caveat #1280: Gitea/git identity from this host may misattribute (issue #1295 showed as created by `@mos-dt-0`); prefer per-invocation `git -c user.name=<seat>` and verify what the remote recorded.
-- Do not delete `sb-it-1-dt:docs]` (untracked file at repo root) — it is cited fleet-wide as incident evidence.
-- Edit user contracts in `~/.mosaic/`, never the templates in `~/.config/mosaic/`.
-- Do not restart other fleet seats unilaterally (goals/scrappy/sanity restart decisions are fred's/Jason's per the docs-seat report).
-
-## 6. Artifact map
-
-| Artifact                                                           | Where                                                                                                                    |
-| ------------------------------------------------------------------ | ------------------------------------------------------------------------------------------------------------------------ |
-| Decision records D-001/D-002 (strict, pending ruling)              | jarvis-brain `docs/decisions/mosaic-stack/2026-08-17_mosaic-d-00{1,2}_*.md`                                              |
-| Incident issue + provisioning comment                              | stack #1295 + comment 23027                                                                                              |
-| ensure-watcher.sh (reference implementation)                       | jarvis-brain `scripts/comms/ensure-watcher.sh` (commit `701c353b1`)                                                      |
-| Bench for prompt A/B (pi, thinking levels, footer-token semantics) | jarvis-brain `domains/software-dev/mosaic-stack/prompt-testing/` + `docs/reports/2026-08-17-prompt-testing-glm-bench.md` |
-| Launcher inspection basis                                          | `@mosaicstack/mosaic` 0.0.49 `dist/commands/launch.js` (composeContract / ensureRuntimeConfig / harness-home isolation)  |
-| Guardrail edits                                                    | `~/.mosaic/{SOUL,AGENTS,STANDARDS}.md` on sb-it-1-dt (2026-08-17 16:53–17:04)                                            |