Gladex Agent Logs

+this section's quote/gloss pair and parks the correct pair in the section
+**after** it, so the next-heading-bounded reader exits 1 while the to-EOF
+mutant reads 0 on the same planted file — the mechanism (125)/`[0.4.182]`
+describes above. The rule for editors of THIS file: **do not rename, reorder,
+split or insert a heading between `## test_registry_coverage.sh` and
+`## test_detached_children.sh` without editing C42's `nxt` literal in
+`tests/test_registry_coverage.sh` in the same change.** Measured cost of
+getting it wrong: run 949 (2026-10-14) inserted
+`## test_registry_figure_readers.sh` between the two and the suite came back
+**488 passed / 9 failed**, all nine reds `C42a`–`C42j` (`C42a` reading
+`want=## test_detached_children.sh got=## test_registry_figure_readers.sh`);
+moving the new section to sit immediately BEFORE this one restored **497
+passed / 0 failed** without touching the suite. The heading after this section
+today is `## test_detached_children.sh` (re-verified at the time of this
+note).
+
+
 ## test_detached_children.sh
 
 **Purpose**: The detached-child sweep of queue item **(80)**, made permanent.
--- heading check ---
5391:## test_registry_coverage.sh
6172:## test_detached_children.sh
--- status ---
 M tools/REGISTRY.md
> build · mimo-v2.6-flash-free
Edit landed cleanly (21 lines, heading order intact). Now verification — launching the long `test_registry_coverage.sh` (497 assertions) in background while running the fast checks:
$ cd /data/repo && bash tests/test_registry_coverage.sh > /tmp/opencode/cov_950.log 2>&1; echo "regcov rc=$?"; tail -3 /tmp/opencode/cov_950.log
Command moved to the background (shell ID: sh_1385d27c4001GaZMa5pXCauRJ2).
Output is streaming to: /root/.local/share/opencode/shell/fa3e85b2145a60c127b5c9261e76a7ba6d520299/sh_1385d27c4001GaZMa5pXCauRJ2.out
$ cd /data/repo && echo '=== figure readers suite ==='; bash tests/test_registry_figure_readers.sh 2>&1 | tail -6; echo "rc=$?"; echo '=== repo-lint (citation watch discharge) ==='; ./tools/repo-lint 2>&1 | tail -3; echo "rc=$?"
=== figure readers suite ===
CENSUS-read:alpha test_registry_coverage.sh 
CENSUS-unread: budget-show disk-show dns-verify domain-availability-check immich-roundtrip ip-drift-check ip-drift-cron pdns-api.py promote-dev-to-prod regression-run repo-lint shift-deploy smtp-relay-probe source-sync-check system-status test_identity_wrapper.sh test_registry_figure_readers.sh
PASS: L1 non-vacuity: at least 2 live sections carry an assertion figure
PASS: L2 test_registry_coverage.sh's own figure IS read (i_reg_pair)

=== Results: 10 passed, 0 failed ===
rc=0
=== repo-lint (citation watch discharge) ===
  changelog-version: 344 changelog version heading(s), 344 unique, 10599 citation(s) checked, 0 missing, 14090 bare token(s) counted
  elf-allowlist: 2 tracked ELF file(s), all shipped
result: all 213 linted file(s) parse clean (exit 0)
rc=0
> build · mimo-v2.6-flash-free
Fast checks green (10/0, repo-lint 0 missing — citation watch discharged). While the long suite runs, re-checking STEP 0 freshness and tree state:
$ cd /data/repo && ./tools/inbox-status | tail -3; echo "rc=$?"; sqlite3 /opt/startup/dev/data/messages.db "SELECT count(*) FROM messages WHERE direction='investor_to_agent' AND read=0;"; sqlite3 /opt/startup/prod/data/messages.db "SELECT count(*) FROM messages WHERE direction='investor_to_agent' AND read=0;"; echo '=== status ==='; git status --porcelain; echo '=== lena timer ==='; systemctl list-timers 'identity-run@lena.timer' --no-pager | head -2; echo '=== date ==='; date '+%Y-%m-%d %H:%M:%S %Z'
INBOX.md        : 80 entries, 80 handled, 0 open

verdict: OK - nothing owed (0 unread, 0 open entries all replied)
rc=0
0
0
=== status ===
 M tools/REGISTRY.md
=== lena timer ===
NEXT                            LEFT LAST                          PASSED UNIT                    ACTIVATES
Wed 2026-10-14 08:05:17 CEST 3h 1min Tue 2026-10-13 18:06:29 CEST 10h ago identity-run@lena.timer identity-run@lena.service
=== date ===
2026-10-14 05:03:51 CEST
> build · mimo-v2.6-flash-free
While the suite runs, taking the remaining standing measurements for my entry:
$ cd /data/repo && echo '=== promote-gates dry-run (standing watch) ==='; timeout 60 ./tools/promote-dev-to-prod --dry-run --format json 2>&1 | python3 -c "import json,sys; d=json.load(sys.stdin); print('ready:', d.get('ready'), 'blocked_by:', d.get('blocked_by'), 'sha:', d.get('sha'))" 2>&1 | head -3; echo '=== red-watch state ==='; cat /data/agent-logs/red-watch.state; echo '=== rev-list ==='; git rev-list --count origin/main..main; git log --oneline -1
=== promote-gates dry-run (standing watch) ===
Traceback (most recent call last):
  File "<string>", line 1, in <module>
  File "/usr/lib/python3.12/json/__init__.py", line 293, in load
=== red-watch state ===
state=red
since=1791893840
failed=2
fails=FAIL A12 system-status verdict - checks in error: failed-units (kind=is key=sysverdict) ;; FAIL A30 no unit is left in systemctl --failed (kind=contains key=sys)
updated=2026-10-14T03:00:01Z
=== rev-list ===
0
1590690 run 949: run 931's recovered line lands as the 109th suite — tests/test_registry_figure_readers.sh measures which REGISTRY assertion figures are actually read (17 figure sections, 1 read, 16 unread) with four hermetic plants and REPORTS the count; new suite section placed before test_registry_coverage.sh so C42's hardcoded successor stays intact; - Live: 108 -> 109; [0.4.339]
> build · mimo-v2.6-flash-free
$ cd /data/repo && timeout 90 ./tools/promote-dev-to-prod --dry-run --format json 2>&1 | head -c 800; echo; echo "rc=$?"
[2026-10-14T03:04:58.612798Z] Starting promotion dev -> prod
GATE OK: VERDICT-20261013T192305Z-work-attribution.md (VERDICT: APPROVE), 34 verdict file(s), age 24920s, all promoted trees (app/src/php, examples/workflows) unchanged since
DEV-SYNC OK: dev tree (/opt/startup/dev) matches the repo — nothing in dev would be dropped by this promotion
LINT OK: repo-lint verdict on HEAD clean (443 file(s) linted, exit 0)
SHIP-TREE OK: promoted trees match HEAD (app/src/php, examples/workflows)
[DRY-RUN] Would run: git status --porcelain (cwd=/data/repo)
[DRY-RUN] Would run: git rev-parse HEAD (cwd=/data/repo)
Promoting commit: 
Copying source from /data/repo/app/src/php to /opt/startup/prod/src/php
[DRY-RUN] Would run: rsync -av --delete /data/repo/app/src/php/ /opt/startup/prod/src/php/ (cwd=/da
rc=0
> build · mimo-v2.6-flash-free
$ cd /data/repo && timeout 90 ./tools/promote-dev-to-prod --dry-run --format json 2>/dev/null | tail -c 600; echo; echo '---'; cat /root/.local/share/opencode/shell/fa3e85b2145a60c127b5c9261e76a7ba6d520299/sh_1385d27c4001GaZMa5pXCauRJ2.out 2>/dev/null | tail -5
up/dev) matches the repo \u2014 nothing in dev would be dropped by this promotion","child":"/data/repo/tools/source-sync-check"},{"name":"commit-lint","status":"ok","code":0,"detail":"repo-lint verdict on HEAD clean (443 file(s) linted, exit 0)","child":"/data/repo/tools/repo-lint"},{"name":"ship-tree","status":"ok","code":0,"detail":"promoted trees match HEAD (app/src/php, examples/workflows)","child":"/usr/bin/git"}],"ready":true,"blocked_by":[],"sha":"1590690","mailbox":{"dir":"mailboxes/reviewer-to-main","newest":"2026-10-13T20:09:38Z","age_seconds":24955},"performed":false,"exit_code":0}

---
> build · mimo-v2.6-flash-free
$ cd /data/repo && grep -n 'PROGRESS bullet\|bullet' tools/queue-source-check | head -12; echo '--- R8 / fail conditions ---'; grep -n 'R8\|exit 1\|rc=1' tools/queue-source-check | head -25
8:    `Next-candidate queued` bullet - two lists, hand-maintained, in two files.
45:        bullet - the authoritative list still exists
64:    R9  the NEWEST `Next-candidate queued` bullet in agent-logs/PROGRESS.md
66:        it (outside fenced code) exists in the repo tree, unless the bullet
106:    named the newest version, so nothing ever read a queue bullet as a CLAIM
109:    the next run reads the bullet, looks for the path, and either rebuilds work
112:      - the NEWEST bullet only. Older bullets are history this tool freezes by
119:        live bullet says `/data/tools/` while describing a stale copy of this
126:      - the marker is read across the WHOLE bullet, as the queue item words it.
127:        KNOWN FALSE POSITIVE, stated rather than papered over: a bullet that
129:        live bullet quotes all three markers because it describes this rule.
134:    A bullet is read as its opening line plus the continuation lines up to the
--- R8 / fail conditions ---
36:Rules (any failure = exit 1, and every one is reported, not just the first):
60:    R8  the version token of the NEWEST `## [x.y.z]` entry is named in
73:Why R8 exists (queue item (109), candidate (a)), measured before it was
89:    file, so R8 is judged only when a newest version exists rather than
105:    judge SHAPE (counts, pointers, headings) and R8 asks only whether PROGRESS
314:    R8's reader, kept separate from `count_queue_bullets()` for the same reason
721:    # R8 - the newest entry recorded itself in the authoritative file. Written
739:    # R9 - the newest bullet's own claims about the tree. Written last with R8
> build · mimo-v2.6-flash-free
$ cd /data/repo && sed -n '56,72p' tools/queue-source-check; echo '--- R8 reader impl ---'; sed -n '714,745p' tools/queue-source-check
written above the first `## [x.y.z]` heading, or a future edit that
        stops appending) is a dropped heading, not a file without one. The two
        readers walk the same file separately precisely so they can disagree;
        one pass reporting both numbers could not.
    R8  the version token of the NEWEST `## [x.y.z]` entry is named in
        agent-logs/PROGRESS.md - outside fenced code, on digit/dot boundaries
        - so the run that wrote the newest CHANGELOG entry is recorded in the
        authoritative file too.
    R9  the NEWEST `Next-candidate queued` bullet in agent-logs/PROGRESS.md
        names no phantom path: every `tools/<name>` / `tests/<name>` token in
        it (outside fenced code) exists in the repo tree, unless the bullet
        marks it new (`new tool` / `new suite` / `(new)`)
    R10 the commit that last touched CHANGELOG.md is an ancestor of, or equal
        to, the commit that last touched agent-logs/PROGRESS.md - the recency
        of the newest changelog entry against the newest record of it, asked
        in COMMIT order (queue item (109), candidate (b), REDESIGNED)

--- R8 reader impl ---
    if raw_headings != accounted:
        violations.append(
            "CHANGELOG.md carries %d '### Queue' heading(s) but %d section(s) "
            "+ %d fenced = %d - a heading was dropped"
            % (raw_headings, len(sections), fenced_headings, accounted)
        )

    # R8 - the newest entry recorded itself in the authoritative file. Written
    # last because it is the rule that reads the two files as ONE story rather
    # than as two shapes: R1-R7 can all be green over a PROGRESS.md that is
    # merely behind CHANGELOG's newest entry, which is what [0.4.160] carried
    # for 23 minutes. Judged only when a newest version exists (with no
    # `## [x.y.z]` heading the newest-entry rules already refuse the file, and
    # naming "the newest entry" of a file that has none would be inventing one).
    newest_ver = parsed["newest_version"]
    if newest_ver:
        named = version_named_outside_fences(pg_text, newest_ver)
        report["rules"]["newest_version_in_progress"] = named
        if not named:
            violations.append(
                "[%s] is not named in %s - the newest CHANGELOG entry has no "
                "record in the authoritative file"
                % (newest_ver, PROGRESS_REL)
            )

    # R9 - the newest bullet's own claims about the tree. Written last with R8
    # because it reads the same file for the same reason (a queue that is
    # structurally perfect can still be WRONG), but it judges a different
    # thing: not "does PROGRESS record the newest run" - "does the live queue
    # head point at files that exist". Judged only when a newest bullet exists:
    # with no bullet R5 already refuses the file, and inventing one to judge
    # would be the guess the A-rules exist to avoid.
> build · mimo-v2.6-flash-free
While the long suite runs, one more independent guard relevant to my edit (`test_leak_figure_readers.sh` — my note contains no figures):
$ cd /data/repo && timeout 240 bash tests/test_leak_figure_readers.sh 2>&1 | tail -4; echo "rc=$?"
===============================
leak-figure-readers suite: 41 passed, 0 failed
===============================
rc=0
> build · mimo-v2.6-flash-free
$ wc -l /tmp/opencode/cov_950.log; tail -3 /tmp/opencode/cov_950.log; date '+%H:%M:%S'
108 /tmp/opencode/cov_950.log
PASS: C21b plant: and the bad VALUE really is swallowed with exit 0
PASS: C21c plant: while a missing value still exits 2 - one class, not two
PASS: C21d plant: and E's probe still passes (this is NOT the (64) defect)
05:07:12
> build · mimo-v2.6-flash-free
The picture is now fully clear (loop cycle 1125 = PROGRESS run 949; its safety-net sweep fired mid-session at 04:33; it finished as `1590690` at 04:56 — overlapping my orientation). Drafting my entry while the suite finishes:
← Write /tmp/opencode/entry_950.md
Created file successfully: /tmp/opencode/entry_950.md
> build · mimo-v2.6-flash-free
$ tail -4 /tmp/opencode/cov_950.log; echo '---'; grep -c '^PASS\|^FAIL' /tmp/opencode/cov_950.log; date '+%H:%M:%S'
PASS: C33a plant: the bullet is present and its date is replaced
PASS: C33b plant: and it still carries a number, so there is still something to be stale
PASS: C33c plant: and it carries no suite figure, so (63)'s selector cannot see it
PASS: C33d plant: and it now has no date anywhere on its line
---
226
05:09:59
> build · mimo-v2.6-flash-free
The 497-assertion suite is at 226 PASS and climbing. My entry draft is ready in `/tmp/opencode/entry_950.md` (working copy only — it will be appended to `agent-logs/PROGRESS.md` **before** the commit, per CLOSE PROTOCOL rule 1). Waiting for the suite's result before appending, since the entry cites its measured outcome.
exit=0

Select another run log from the list above. Only files matching run-YYYYMMDD-HHMMSS-N.log are readable.

[Wed Oct 14 05:36:56 2026] 127.0.0.1:34738 Accepted
[Wed Oct 14 05:36:56 2026] 127.0.0.1:34738 Closing
[Wed Oct 14 05:41:33 2026] 127.0.0.1:54516 Accepted
[Wed Oct 14 05:41:33 2026] 127.0.0.1:54516 Closed without sending a request; it was probably just an unused speculative preconnection
[Wed Oct 14 05:41:33 2026] 127.0.0.1:54516 Closing
[Wed Oct 14 05:41:35 2026] 127.0.0.1:53092 Accepted
[Wed Oct 14 05:41:35 2026] 127.0.0.1:53092 Closed without sending a request; it was probably just an unused speculative preconnection
[Wed Oct 14 05:41:35 2026] 127.0.0.1:53092 Closing
[Wed Oct 14 05:42:23 2026] 127.0.0.1:39072 Accepted
[Wed Oct 14 05:42:23 2026] 127.0.0.1:39072 Closed without sending a request; it was probably just an unused speculative preconnection
[Wed Oct 14 05:42:23 2026] 127.0.0.1:39072 Closing
[Wed Oct 14 05:42:24 2026] 127.0.0.1:39076 Accepted
[Wed Oct 14 05:42:24 2026] 127.0.0.1:39076 Closed without sending a request; it was probably just an unused speculative preconnection
[Wed Oct 14 05:42:24 2026] 127.0.0.1:39076 Closing
[Wed Oct 14 05:45:22 2026] 127.0.0.1:52848 Accepted
[Wed Oct 14 05:45:22 2026] 127.0.0.1:52848 Closing
[Wed Oct 14 05:45:22 2026] 127.0.0.1:52856 Accepted
[Wed Oct 14 05:45:22 2026] 127.0.0.1:52856 Closing
[Wed Oct 14 05:45:22 2026] 127.0.0.1:52862 Accepted
[Wed Oct 14 05:45:22 2026] 127.0.0.1:52862 Closing
[Wed Oct 14 05:45:22 2026] 127.0.0.1:52866 Accepted
[Wed Oct 14 05:45:22 2026] 127.0.0.1:52866 Closing
[Wed Oct 14 05:45:22 2026] 127.0.0.1:52868 Accepted
[Wed Oct 14 05:45:22 2026] 127.0.0.1:52868 Closing
[Wed Oct 14 05:45:22 2026] 127.0.0.1:52882 Accepted
[Wed Oct 14 05:45:22 2026] 127.0.0.1:52882 Closing
[Wed Oct 14 05:45:22 2026] 127.0.0.1:52892 Accepted
[Wed Oct 14 05:45:22 2026] 127.0.0.1:52892 Closing
[Wed Oct 14 05:45:22 2026] 127.0.0.1:52908 Accepted
[Wed Oct 14 05:45:22 2026] 127.0.0.1:52908 Closing
[Wed Oct 14 05:45:22 2026] 127.0.0.1:52914 Accepted
[Wed Oct 14 05:45:23 2026] 127.0.0.1:52914 Closing
[Wed Oct 14 05:45:23 2026] 127.0.0.1:52926 Accepted
[Wed Oct 14 05:45:23 2026] 127.0.0.1:52926 Closing
[Wed Oct 14 05:46:28 2026] 127.0.0.1:44264 Accepted
[Wed Oct 14 05:46:28 2026] 127.0.0.1:44264 Closing
[Wed Oct 14 05:46:28 2026] 127.0.0.1:44272 Accepted
[Wed Oct 14 05:46:28 2026] 127.0.0.1:44272 Closing
[Wed Oct 14 05:46:28 2026] 127.0.0.1:44280 Accepted
[Wed Oct 14 05:46:28 2026] 127.0.0.1:44280 Closing
[Wed Oct 14 05:46:28 2026] 127.0.0.1:44284 Accepted
[Wed Oct 14 05:46:28 2026] 127.0.0.1:44284 Closing
[Wed Oct 14 05:46:28 2026] 127.0.0.1:44290 Accepted
[Wed Oct 14 05:46:28 2026] 127.0.0.1:44290 Closing
[Wed Oct 14 05:46:28 2026] 127.0.0.1:44292 Accepted
[Wed Oct 14 05:46:28 2026] 127.0.0.1:44292 Closing
[Wed Oct 14 05:46:28 2026] 127.0.0.1:44308 Accepted
[Wed Oct 14 05:46:29 2026] 127.0.0.1:44308 Closing
[Wed Oct 14 05:46:29 2026] 127.0.0.1:44316 Accepted
[Wed Oct 14 05:46:29 2026] 127.0.0.1:44316 Closing
[Wed Oct 14 05:46:29 2026] 127.0.0.1:44320 Accepted
[Wed Oct 14 05:46:29 2026] 127.0.0.1:44320 Closing
[Wed Oct 14 05:46:29 2026] 127.0.0.1:44336 Accepted
[Wed Oct 14 05:46:29 2026] 127.0.0.1:44336 Closing
[Wed Oct 14 05:48:58 2026] 127.0.0.1:36620 Accepted
[Wed Oct 14 05:48:58 2026] 127.0.0.1:36620 Closing
[Wed Oct 14 05:49:19 2026] 127.0.0.1:42978 Accepted
[Wed Oct 14 05:49:19 2026] 127.0.0.1:42978 Closing
[Wed Oct 14 05:49:19 2026] 127.0.0.1:42988 Accepted

Other app logs: downloads.log