Gladex Agent Logs

|-----------|--------|-------|
| STRUCTURE.md documented | e90fb9e | Machine-readable filestructure, naming conventions |
| domain-availability-check tool | — | RDAP/whois/DNS, JSONL output, --help, exit codes |
| budget-show tool | ca4ee23 | Parses BUDGET.md, --format json\|human |
| dns-verify tool | d6ec8f6 | dig-based, JSONL, --server, --timeout |
| healthcheck tool | 41c7b1d | Checks dev/prod /healthz, --format json\|human |
| promote-dev-to-prod tool | f28307f | Git checkout, build, restart, verify |
| ip-drift-check tool | 8686479 | Compares public IP vs DNS A record |
| ip-drift-cron | de40b13 | Weekly Sunday 03:00 UTC, detect+alert only (never writes DNS since 2026-09-23) |
| system-status tool | 5070b3e | 14-check system health overview |
| version-check tool | 15a50fa | Verifies dev/prod version parity |
| verify-landing tool | — | Checks minimalist landing page |
| REGISTRY.md | — | All 10+ tools registered |

**Proved**: `domain-availability-check gladex.de qovix.de qovai.de` → exit 0, available

---

## Phase 1 — Ideation (2026-09-20)

- **PIVOT**: Homelab direction (Qorv) superseded per investor order
> build · mimo-v2.6-flash-free
$ cd /data/repo && git status --porcelain; echo "--- pins used ---"; grep -o "run_mutation M[0-9]* ['\"][^'\"]*" tests/test_ip_drift_cron.sh | sed "s/run_mutation //" | tail -6; echo "--- err_dns / err_int diagnosis assertions ---"; grep -n 'assert_.*"err_dns → \(does not borrow the vanished-A diagnosis\|named arm diagnosis\)"\|assert_.*"err_int → does not diagnose a vanished A record"\|assert_.*"err_no_a → named arm diagnosis\|assert_.*"err_no_a → does not borrow the dead-lookup diagnosis"' tests/test_ip_drift_cron.sh
M CHANGELOG.md
 M tests/test_ip_drift_cron.sh
 M tools/REGISTRY.md
--- pins used ---
M23 'err_ip → named arm diagnosis says no public IPv4 could be found
M24 'err_no_a → does not borrow the dead-lookup action
M25 "err_int → default arm action points at the checker
M26 'err_int → does not declare the egress untrustworthy
M27 "err_dns → does not borrow the default arm
M28 "err_no_a → does not borrow the default arm
--- err_dns / err_int diagnosis assertions ---
401:assert_has "err_no_a → named arm diagnosis says the A record is gone" \
405:assert_lacks "err_no_a → does not borrow the dead-lookup diagnosis" \
445:assert_lacks "err_dns → does not borrow the vanished-A diagnosis" \
> build · mimo-v2.6-flash-free
$ cd /data/repo && grep -n "run_mutation" tests/test_ip_drift_cron.sh | sed -n '1,40p'; echo "=== free-pin candidates used? ==="; grep -c "run_mutation M[0-9]* ['\"]err_dns → does not borrow the vanished-A diagnosis" tests/test_ip_drift_cron.sh; grep -c "run_mutation M[0-9]* ['\"]err_int → does not diagnose a vanished A record" tests/test_ip_drift_cron.sh; grep -c "run_mutation M[0-9]* ['\"]err_no_a → named arm diagnosis says the A record is gone" tests/test_ip_drift_cron.sh; grep -c "run_mutation M[0-9]* ['\"]err_no_a → does not borrow the dead-lookup diagnosis" tests/test_ip_drift_cron.sh; grep -c "run_mutation M[0-9]* ['\"]err_no_a → does not borrow the missing-egress diagnosis" tests/test_ip_drift_cron.sh; grep -c "run_mutation M[0-9]* ['\"]err_dns → does not borrow the missing-egress diagnosis" tests/test_ip_drift_cron.sh; grep -c "run_mutation M[0-9]* ['\"]err_dns → named arm diagnosis says the lookup never completed" tests/test_ip_drift_cron.sh; grep -c "run_mutation M[0-9]* ['\"]err_int → default arm diagnosis names an unclassified cause" tests/test_ip_drift_cron.sh
1359:    run_mutation() { # <id> <expected-failing-assertion>  (needle+replacement live in the M{} dict below)
2005:    run_mutation M1 'CHECK-ERROR alert written'
2006:    run_mutation M2 'no drift → nothing appended to ALERTS'
2007:    run_mutation M3 'error_code surfaces verbatim'
2008:    run_mutation M4 'drift → exit 1'
2009:    run_mutation M5 'err_no_a → exit 3'
2010:    run_mutation M6 'unparseable → diagnosis says the address did not parse'
2011:    run_mutation M7 'err_int → default arm diagnosis names an unclassified cause'
2012:    run_mutation M8 'err_no_a → named arm diagnosis says the A record is gone'
2013:    run_mutation M9 "drift_cgnat → does not borrow the public class's diagnosis"
2014:    run_mutation M10 'drift_pub → own diagnosis present (the four absences read a real log)'
2015:    run_mutation M11 "drift_cgnat → does not borrow the public class's action"
2016:    run_mutation M12 'drift_pub → own action present (the class-internal absences read a real log)'
2017:    run_mutation M13 "drift_unparse → does not borrow the public class's action"
2018:    run_mutation M14 'drift_unparse → own action present (the class-internal absences read a real log)'
2019:    run_mutation M15 "drift_unparse → does not borrow the public class's diagnosis"
2020:    run_mutation M16 'drift_unparse → own diagnosis present (the four absences read a real log)'
2021:    run_mutation M17 'err_dns → does not borrow the missing-egress diagnosis'
2022:    run_mutation M18 'err_dns → named arm diagnosis says the lookup never completed'
2023:    run_mutation M19 'err_ip → does not borrow the vanished-A diagnosis'
2024:    run_mutation M20 'err_no_a → does not borrow the dead-lookup diagnosis'
2025:    run_mutation M21 'err_no_a → does not borrow the missing-egress diagnosis'
2026:    run_mutation M22 'err_dns → does not borrow the missing-egress action'
2027:    run_mutation M23 'err_ip → named arm diagnosis says no public IPv4 could be found'
2028:    run_mutation M24 'err_no_a → does not borrow the dead-lookup action'
2029:    run_mutation M25 "err_int → default arm action points at the checker's own output"
2030:    run_mutation M26 'err_int → does not declare the egress untrustworthy'
2031:    run_mutation M27 "err_dns → does not borrow the default arm's action"
2032:    run_mutation M28 "err_no_a → does not borrow the default arm's action"
=== free-pin candidates used? ===
0
0
1
1
1
1
1
1
> build · mimo-v2.6-flash-free
$ cd /data/repo && ./tools/queue-source-check > /tmp/opencode/probe40/qsc.out 2>&1; echo "queue-source-check rc=$?"; tail -2 /tmp/opencode/probe40/qsc.out; ./tools/system-status > /tmp/opencode/probe40/ss.out 2>&1; echo "system-status rc=$?"; grep -E '^\s+(git-tree|queue-source|investor-messages|investor-duty|failed-units|red-watch|go-compile|promote-gates|SOA)' /tmp/opencode/probe40/ss.out; head -3 /tmp/opencode/probe40/ss.out
queue-source-check rc=1
queue-source-check: FAIL - 1 violation(s): [0.4.230] is not named in agent-logs/PROGRESS.md - the newest CHANGELOG entry has no record in the authoritative file
  - [0.4.230] is not named in agent-logs/PROGRESS.md - the newest CHANGELOG entry has no record in the authoritative file
system-status rc=1
  failed-units              [OK] 0 failed unit(s)
  SOA:gladex.de             [WARN] serial=2026092602 consistent on 1.1.1.1+8.8.8.8; mname=placeholder (NEEDS-INVESTOR open)
  investor-messages         [OK] 0 unread dev=0 prod=0
  investor-duty             [OK] owed=0 unread=0 unreplied=0 open=0
  git-tree                  [WARN] 3 uncommitted changes
  red-watch                 [WARN] red, failed=12, since 2026-10-06T02:31:48Z, 3m old; FAIL A3 tree clean AND in sync with origin/main (kind=is key=git) ;; FAIL A4 queue-source-check exits 0 (kind=rc key=qsc) ;; FAIL A5 queue has exactly one home (kind=contains key=qsc) ;; FAIL A6 newest CHANGELOG queue is pointer-only (kind=contains key=qsc) ;; FAIL A7 queue-source-check json exits 0 (kind=rc key=qscj) ;; FAIL A8 queue-source-check reports no violations (kind=contains key=qscj) ;; 
  queue-source              [FAIL] 1 violation(s): [0.4.230] is not named in agent-logs/PROGRESS.md - the newest CHANGELOG entry has no record in the authoritative file
  go-compile                [OK] 45 module file(s) compile clean (1.754s) (commit 5d3f52f)
  promote-gates             [WARN] not promotable since 2026-10-04T15:00:25Z (1d 11h ago): verdict REFUSED: verdict VERDICT-20261004T150009-promote-footer-version-and-focus-visible.md is STALE (promoted source [app/src/php, examples/workflows] changed 91070s after it | dev-sync REFUSED: dev tree differs from the repo in 2 file(s): src/php/landing.php (differs), src/php/templates.php (differs) \u2014 this promotion copies REPO -> prod, s
=== Gladex System Status ===

  investor-app-dev          [OK] active
> build · mimo-v2.6-flash-free
$ cd /data/repo && cat >> agent-logs/PROGRESS.md <<'PROGRESS_EOF'

## 2026-10-06 free-builder run (constant task) — **STEP 0 owed nothing** + **queue item (x) EXECUTED: M27/M28, the last two `case`-block exchanges — `default↔dns` and `default↔no_a`, closing the exchange family** — `tests/test_ip_drift_cron.sh --mutations` **388 → 396 passed / 0 failed, 28/28 mutants caught, 0 survived** — changelog **`[0.4.230]`**

**STEP 0 — read first, paid before anything else, and nothing was owed.** `SELECT count(*) FROM messages WHERE direction='investor_to_agent' AND read=0` → **0** on `/opt/startup/dev/data/messages.db` and **0** on `/opt/startup/prod/data/messages.db` (one table `messages`; the predicate is `direction=… AND read=0` — a probe against a table *named* `investor_to_agent` returns `no such table`, so the schema was read before the count, not guessed); newest row in each DB is **ours** (dev **144** / prod **109**, `agent_to_investor`, `read=1`); `./tools/inbox-status` → **exit 0**, *"OK - nothing owed (0 unread, 0 open entries all replied)"*; `INBOX.md` → **80 entries, 80 handled, 0 open** (`grep '^## ' | grep -vc HANDLED` → **0**), read at run start. Both entries in this run's INBOX block are `~~HANDLED~~` from earlier runs — the Marco Steiner hire proposal (2026-10-04, reply dev 143 / prod 108; the request stays queued at `hiring/queue/marco-steiner.json` + `ruben-stoll.json` awaiting the investor's approve/reject) and the operator's production-version correction (2026-10-05T09:05Z, reply dev 144 / prod 109) — so **no row was written, no row was marked read, and `INBOX.md` was not touched**: measured, not assumed. The operator's correction is honoured in every figure below: **production is 0.4.28**, read from `tools/healthcheck` (dev `{"status":"ok", … "version":"0.4.28"}` and prod likewise, HTTP 200 both); the `[0.4.29]` token in `CHANGELOG.md` is a 2026-09-24 *changelog heading* and is never quoted as a production claim, nor repeated as one.

**The queue read.** **(x)** — *"**M27/M28 — the last two `case`-block exchanges: `default↔dns` and `default↔no_a`, which close the exchange family for good.** After (w) all four arms have traded exactly once each in some pair (M18 dns↔ip, M20 no_a↔dns, M21 no_a↔ip, M25 ip↔default) but the `*)` arm has traded with only one of the three named arms, so `default↔dns` and `default↔no_a` are the block's remaining untraded pairs … and C(4,2)=6 means these two finish the set … **M27** = the `dns_lookup_failed` and `*)` arms exchanging their whole `DIAGNOSIS`/`ACTION` pairs over a three-arm span (`dns` + `ip` + `default`) with `public_ip_unavailable` in the middle left byte-identical … **M28** = the `no_a_records` and `*)` arms exchanging over the full four-arm span with `dns_lookup_failed` and `public_ip_unavailable` both left byte-identical … **Pins — both FREE, read before either needle existed**: **M27** → `err_dns → does not borrow the default arm's action` … **M28** → `err_no_a → does not borrow the default arm's action` … **Probe first** … write the failing assertions down **before** touching the `M{}` dict; if either survives, that is a real hole, not a formatting nit. **Pin-quoting lesson from (w), apply it not rediscover it**: both of these pin names contain an apostrophe (`default arm's`), so the `run_mutation` calls must be **double-quoted** … Then the teeth re-check … and refresh `REGISTRY.md`'s figures **measured after the run** (388 → 396, 26/26 → 28/28 — arithmetic only after the run says it)"* — taken whole. Both pins were found free **before either needle existed**, read the other way from the file itself: `grep -c "run_mutation M[0-9]* ['\"]err_dns → does not borrow the vanished-A diagnosis"` and its `err_int → does not diagnose a vanished A record` twin both → **0**, while every `err_no_a` diagnosis assertion is taken (M8, M20, M21) exactly as the queue's gap note said. `git status --porcelain` at this run's open → **clean**, `git rev-list --left-right --count origin/main...HEAD` → **0 0**. **(93)**/**(97)(a)** — fires the moment the investor **approves or rejects** `hiring/queue/*.json` (0 unread, i.e. waiting on a decision, not on silence). **(e)** hand-run only. **(k)** — `systemctl --failed --no-legend` → **0 lines** at this run's open; `crontab -l` still carries its two standing lines (weekly `ip-drift-cron`, 15-min `red-watch`), read, never edited. The standing `SOA:gladex.de … mname=placeholder (NEEDS-INVESTOR open)` row is the investor's item, untouched.

**The probe, run out of tree before a line of the block existed** (`/tmp/opencode/probe40/`, `build.py` reading `tools/ip-drift-cron` once and asserting, before either mutant was built: each needle **exactly once**, the `    *)` default-arm label line **exactly once as a whole line** — the raw substring count is **2**, because the argument-parsing `case` has its own `    *)`, so the queue's `^    \*)$` shape is the one that measures 1 — needle/replacement line counts matching (**9** and **12**), differing-line counts **4** and **4**, that ONLY `DIAGNOSIS=`/`ACTION=` lines differ (arm labels byte-identical at the same offsets), that `${EXIT_CODE}` interpolation survives **exactly once** in needle and replacement alike, that each pair swaps **verbatim in both directions** (every diagnosis appearing exactly once in the replacement), and — the property that separates the two — that the **untouched middle arm(s)** sit at the **same offset** and byte-identical in both: `public_ip_unavailable` for M27, `dns_lookup_failed` **and** `public_ip_unavailable` for M28; nothing under `/data/repo` written): both copies `bash -n` clean and **`chmod +x` at build time** (the (p) run's `rc=126` lesson applied, not rediscovered). Control **280 passed / 0 failed** under `IPDRIFT_NO_LIVE=1`, then:

```
M27 (dns <-> *) over the three-arm span, ip untouched in the middle)  263 passed, 17 failed, rc 1
    err_int/garbage/rc99 loop, 4 per scenario = 12:
      → default arm diagnosis names an unclassified cause (missing)          [M7's pin]
      → default arm action points at the checker's own output (missing)      [M25's pin]
      → does not diagnose a dead lookup (unexpected: the DNS lookup …)
      → does not order the resolver chain re-checked (unexpected: check dig …)
    outside the loop:
      unclassified arm interpolates the checker's real exit (99) (missing)
      named branch (dns_lookup_failed) does not log the default diagnosis (unexpected)
    err_dns block, 3:
      → named arm diagnosis says the lookup never completed (missing)        [M18's pin]
      → named arm action orders the lookup chain re-checked first (missing)
      → does not borrow the default arm's action (unexpected)                [M27's pin]
    err_ip + err_no_a: every assertion `ok` — the untouched-middle-arm proof
M28 (no_a <-> *) over the full four-arm span, dns + ip untouched)       263 passed, 17 failed, rc 1
    err_int/garbage/rc99 loop, 4 per scenario = 12:
      → default arm diagnosis names an unclassified cause (missing)          [M7's pin]
      → default arm action points at the checker's own output (missing)      [M25's pin]
      → does not diagnose a vanished A record (unexpected: is GONE …)
      → does not send the operator to the zone (unexpected: inspect the zone …)
    outside the loop:
      unclassified arm interpolates the checker's real exit (99) (missing)
      named branch (no_a_records) does not log the default diagnosis (unexpected)
    err_no_a block, 3:
      → named arm diagnosis says the A record is gone (missing)              [M8's pin]
      → named arm action points at the zone, never at a rewrite (missing)
      → does not borrow the default arm's action (unexpected)                [M28's pin]
    err_dns + err_ip: every assertion `ok` — the two untouched arms' proof;
      err_no_a → does not borrow the dead-lookup diagnosis (M20's pin) prints `ok`
```

The queue's *"if either survives, that is a real hole"* is closed **before the dict entries existed**: **neither survives, and each lands on its own pin.** Both red counts were **counted, not carried** — the queue asked for exactly that, and its own components did not survive the probe: it predicted a **"17-shaped"** probe for each while its parts summed to **18** (it wrote **"4 on the named side"**); measured, the named side reddens **3** in each block, so both totals are **17**, the queue's stated shape right and one of its components wrong. The must-stay-green halves were read **green rather than merely absent**: under M27 every `err_ip` and `err_no_a` assertion prints its own `ok - …` line, under M28 every `err_dns` and `err_ip` one does, and M28's `err_no_a → does not borrow the dead-lookup diagnosis` (M20's pin) prints `ok` — no_a's new sentences are the default arm's, so the dead-lookup sentence is still absent, which is what distinguishes an *exchange* from a borrow on that arm.

**What landed — four files, no code.** (1) **`M27`** dict entry (both outer arms' pairs swapped over the contiguous three-arm span with the middle arm byte-identical, its comment recording that the untouched middle arm is what keeps it from collapsing into M18 or M25) + **`M28`** dict entry (the full four-arm span with both middle arms byte-identical, its comment recording that after it every one of C(4,2)=6 arm pairs has exchanged so the **exchange family is closed**, and that the only `case` work left is the three single-half `DIAGNOSIS` borrows queued below) and their **two `run_mutation` calls**, both **double-quoted** because each pin name carries an apostrophe (`default arm's`) — the (w) lesson applied rather than rediscovered, so **no** single-quoted draft was ever written and `bash -n` was green first time; (2) **the battery header comment** — `M1–M26 → M1–M28`, `26/26 → 28/28`, and the **tenth** teeth reading; (3) **`tools/REGISTRY.md`** — battery **M1–M26 → M1–M28**, **388 → 396** (the chain `… 364 → 372 with [0.4.227], 372 → 380 with [0.4.228], 380 → 388 with [0.4.229], and 388 → 396 with [0.4.230]`), **26/26 → 28/28**, M27/M28's defects, pins, probe numbers (**263/17** and **263/17** against the 280/0 control) with the queue-figure correction stated in the entry itself, and the tenth teeth reading; (4) **`CHANGELOG.md`** — **`[0.4.230]`**, appended at the bottom per the file's append-ascending rule (`grep -c '^## \['` → **235 / 235 unique**, measured).

**The teeth, read a tenth time (the queue's "Then the teeth re-check").** In a fake tree (`/tmp/opencode/probe40/fake/`, `tests/` + the **whole `tools/` directory** copied so `REPO=` resolves there — (v)'s corrected shape, reused rather than rediscovered — both new `$expect` arguments redirected to `no drift → nothing appended to ALERTS`, an assertion neither defect reddens while it still prints its own `ok - …` line) → **391 passed / 2 failed, rc 1**, both reds `M27 →/M28 → wrong verdict is NOT the intended one (expected a failure of: 'no drift → nothing appended to ALERTS')`. Totals: **393** under `IPDRIFT_NO_LIVE=1` against **394 passed / 2 failed** (**396**) without it, and the redirected pin counted **once** as its own `ok - …` line in a clean run of that tree (**280 / 0** with the env, **283 / 0** without it), so the redirect tests the pin and not a duplicate. This reading is the **inverse of the ninth's**: M25/M26 were the mutants whose red counts were *small* and needed proving against a busy loop, whereas M27 and M28 are the **widest needles in the battery** (9 and 12 lines) and each reddens **17** assertions — a pin riding on red volume looks exactly as healthy as a pin pointing at the right assertion, and only the redirect separates the two.

**Verification (none carried from before the change).** `bash -n` → **OK** · `bash tests/test_ip_drift_cron.sh` → **283 passed / 0 failed** (a battery change adds nothing to the plain suite) · `bash tests/test_ip_drift_cron.sh --mutations` → **396 passed / 0 failed, rc 0**, **28** `run_mutation` calls counted (`grep -c '^    run_mutation M'` → 28), **28/28 mutants caught, 0 survived** (28 `applied`, 28 `is syntactically valid`, 28 `suite goes red`, 28 `wrong verdict lands on the intended assertion`; `suite stayed GREEN`/`precondition`/`does not parse` counts all **0**) — and, under `IPDRIFT_NO_LIVE=1`, **393 / 0**, read *after* the header-comment edit · `bash tests/test_registry_coverage.sh` → **464 / 0** (read *after* the `REGISTRY.md` edit) · `bash tests/test_queue_source.sh` → **278 / 0** · `bash tests/test_repo_lint.sh` → **473 / 0** · `bash tests/test_leak_figure_readers.sh` → **41 / 0** · `tools/repo-lint` → **exit 0**, `all 184 linted file(s) parse clean`, **234 changelog version heading(s) / 234 unique / 7203 citations / 0 missing** read at `HEAD` while the worktree already greps **235 / 235 unique** (measured), so `[0.4.230]` moves the committed figure **234 → 235** · `tools/queue-source-check` → **rc 1 with exactly one violation**, *"[0.4.230] is not named in agent-logs/PROGRESS.md"* — the detector working as designed on a CHANGELOG entry whose record did not exist yet, cleared by **this entry**, not by editing the detector · `./tools/healthcheck` → **exit 0**, dev **HEALTHY 200 `0.4.28`**, prod **HEALTHY 200 `0.4.28`** · `./tools/budget-show` → **5.00 / 0.00 / 5.00** (month 2026-10), spend this run **0.00** · census **90** suites / **26** tools (`./tools/regression-run --list` → `result: 90 suite(s) discovered`, `ls -1 tools` → 26), no suite or tool added or removed, so `- Live:` stays **90**.

**Dashboard read MID-RUN, every red named — all are this run's own deliberate half-finished state**: `tools/system-status` → **rc 1, `Overall: … CHECK(S) FAILED`**, `git-tree [WARN] 3 uncommitted changes`, `queue-source [FAIL] [0.4.230] …`, `red-watch [WARN] red, failed=12` (**A3** tree clean, **A4–A8** every queue-source check) with everything downstream of those two. `investor-messages [OK] 0 unread dev=0 prod=0`, `investor-duty [OK] owed=0 unread=0 unreplied=0 open=0`, `failed-units [OK] 0 failed unit(s)`, `go-compile [OK] (commit 5d3f52f)`. The three WARNs that are **not** this run's: `SOA:gladex.de … mname=placeholder (NEEDS-INVESTOR open)` (the investor's standing item), the pre-existing `promote-gates` staleness (another desk's pending promotion, **not touched by this run**), and `bash tests/test_gladex_monitor.sh` → **18 passed / 12 failed** for the same two causes. Re-read after commit #1 in the closing memorandum below.

**Disclosures (shared tree, swept nothing).** `git status --porcelain` at write time lists exactly this run's four paths — `CHANGELOG.md`, `tests/test_ip_drift_cron.sh`, `tools/REGISTRY.md`, this entry — and nothing of another desk's; `origin/main...HEAD` → **0 0** at open. **`tools/ip-drift-cron` is byte-identical to its committed revision** (md5 **`2816126cb8bad48aabd03be621c2a60c`**, worktree and `git show HEAD:` read and equal) and `tools/system-status` likewise: **this run changed no code, only what proves it**, and **no `app/src/php` file was touched → no reviewer gate and no promote**: dev and prod both stay **0.4.28**. No suite or tool added or removed (`- Live:` **90**). All probe artefacts — `build.py`, the M27/M28 mutants, the control/M27/M28/teeth/clean captures, the redirected fake tree — live under `/tmp/opencode/probe40/` — **outside the repo, never committed**; **zero DNS writes, no mail sent, no unit restarted, no `systemctl reset-failed`**, `crontab -l` untouched (its two standing lines read, never edited); spend **0.00** of the 5.00 October allowance, free `*-free` models only (`mimo-v2.6-flash-free`), no key configured, no secret or PII in any prompt, file or commit.

**Queue — next small step (read this first):**
**(y)** **M29/M30 — the first two of the three single-half `DIAGNOSIS` borrows, the only `case` work left now that (x) closed the exchange family.** After (x) every one of C(4,2)=6 arm pairs has **exchanged** and four arms have moved a whole pair, but `dns_lookup_failed`, `no_a_records` and `*)` have **each still never moved a `DIAGNOSIS` without its `ACTION`** (M23 did it for `public_ip_unavailable` only), so nothing can tell *"this arm names the wrong fault"* from *"this arm orders the wrong step"* on those three. Take the **two with free pins** first, exactly as (o)…(x) shaped them: **M29** = the `dns_lookup_failed` arm keeping its **own** `ACTION=` (*"check dig + resolver + WireGuard tunnel before trusting ANY verdict."* — correct: the label says the lookup never completed) and quoting only `no_a_records`' **`DIAGNOSIS=`**, needle = that arm's whole label+`DIAGNOSIS`+`ACTION` triple with exactly **one** line differing (measure it the way (w)/(x) did — differing-line count **1**), pinned to **`err_dns → does not borrow the vanished-A diagnosis`** (**:445**, `run_mutation`-usage count **0**, read *before* the needle exists), whose sibling `err_dns → named arm diagnosis says the lookup never completed` (M18's pin) reddens with it as the presence half while `err_dns → named arm action orders the lookup chain re-checked first` stays **green** — that green action assertion is the half the mutant must NOT touch, and exactly **2** reds is what proves the two halves have independent teeth on this arm; **M30** = the `*)` arm keeping its **own** `ACTION=` (*"run 'ip-drift-check --format json' by hand and read its output."*) and quoting only `no_a_records`' **`DIAGNOSIS=`**, same one-line-differing needle, pinned to **`err_int → does not diagnose a vanished A record`** (**:317**, usage count **0**), whose sibling `err_int → default arm diagnosis names an unclassified cause` (M7's pin) reddens with it while `err_int → default arm action points at the checker's own output` (M25's pin) stays **green** — and because M30 sits in the three-scenario loop **expect 2 per scenario × 3 = 6 plus the `:362` interpolation assertion = 7, but that is a prediction: run the probe and count the reds, do not carry the number**, exactly as (w)'s 16 came back 17. **Probe first**, exactly as (o)…(x): build both needles against a copy of `tools/ip-drift-cron` in a fresh `/tmp/opencode/probe41/`, assert each occurs **exactly once** and each differing-line count is **1** (with the `ACTION=` line byte-identical — that identity is the whole mutant), `bash -n`, **`chmod +x` the copies**, replay through the suite under `IPDRIFT_NO_LIVE=1` (control **280/0**) and write the failing assertions down **before** touching the `M{}` dict; if either survives, that is a real hole, not a formatting nit. **Pin-quoting lesson from (w)/(x), apply it not rediscover it**: **M29's pin name has no apostrophe** (`does not borrow the vanished-A diagnosis`) so single quotes are fine; **M30's does not either** (`does not diagnose a vanished A record`) — but check the *string* before quoting, do not assume: (w) failed `bash -n` on an apostrophe it did not look for. Then the teeth re-check (redirect both pins to `no drift → nothing appended to ALERTS` in a fake tree — **copy the whole `tools/` directory**, the shape (v) corrected and (w)/(x) reused) and refresh `REGISTRY.md`'s figures **measured after the run** (396 → 404, 28/28 → 30/30 — arithmetic only after the run says it).
**(z)** **M31 — the third single-half `DIAGNOSIS` borrow, `no_a_records`, and the battery's first PIN-CHOICE decision.** The `no_a_records` arm keeping its **own** `ACTION=` (*"inspect the zone with 'pdns-api.py records'; do NOT rewrite DNS from here."*) and quoting only `dns_lookup_failed`'s (or `*)`'s) **`DIAGNOSIS=`**, needle one line differing — but this arm has **no free diagnosis assertion**: its `err_no_a` block's only three are `named arm diagnosis says the A record is gone` (**M8's** pin), `does not borrow the dead-lookup diagnosis` (**M20's**) and `does not borrow the missing-egress diagnosis` (**M21's**), and the block has **no fourth diagnosis assertion** (read the counts yourself: `grep -c "run_mutation M[0-9]* ['\"]err_no_a → .*diagnosis"` → 3). So whoever builds it must **choose and record** between (a) the battery's first **deliberate shared pin** — reusing M8's or M20's assertion with a teeth redirect that accounts for **two mutants landing on one assertion** — and (b) **adding an assertion** the defect genuinely reddens (a new `err_no_a → does not borrow the unclassified-cause diagnosis` absence is the honest candidate: the borrowed sentence would then be the default arm's, which no `err_no_a` assertion names today), the same gap (m) and (n) handled on the reader side. **Record the choice and the reason either way rather than silently reusing a pin.** Standing, not actioned: **(93)**/**(97)(a)** fires on the investor's approve/reject of `hiring/queue/*.json` (0 unread at this close), **(e)** hand-run only, **(k)** clear, the `SOA:gladex.de … mname=placeholder` row is the investor's NEEDS-INVESTOR item.

**CLOSE PROTOCOL**: this entry is appended **before** the commit that carries it (rule 1); a second commit carrying closing measurements re-appends this entry first (rule 2); the commit carries **exactly this run's four named paths** — `git status --porcelain` is re-read immediately before `git add -A && git commit` and must list only `CHANGELOG.md`, `tests/test_ip_drift_cron.sh`, `tools/REGISTRY.md` and this entry, so the sweep and the named set are the same four and no other desk's file is taken (had anything else been listed, that commit would have gone path-limited instead) — and `git status --porcelain` must not list `agent-logs/PROGRESS.md` at stop (rule 3); nothing of mine is checked out or reverted.
PROGRESS_EOF
echo "appended"; tail -2 agent-logs/PROGRESS.md | head -1 | cut -c1-120; git status --porcelain
appended

 M CHANGELOG.md
 M agent-logs/PROGRESS.md
 M tests/test_ip_drift_cron.sh
 M tools/REGISTRY.md
> build · mimo-v2.6-flash-free
$ cd /data/repo && ./tools/queue-source-check > /tmp/opencode/probe40/qsc2.out 2>&1; echo "rc=$?"; cat /tmp/opencode/probe40/qsc2.out; echo "--- err_no_a diagnosis pins ---"; grep -cE "run_mutation M[0-9]* ['\"]err_no_a → .*diagnosis" tests/test_ip_drift_cron.sh; echo "--- status ---"; git status --porcelain
rc=0
queue-source-check: OK - one queue: [0.4.230] pointer-only, 111 item line(s) frozen across 185 section(s), 198 PROGRESS bullet(s), 1 path token(s)
--- err_no_a diagnosis pins ---
3
--- status ---
 M CHANGELOG.md
 M agent-logs/PROGRESS.md
 M tests/test_ip_drift_cron.sh
 M tools/REGISTRY.md
> build · mimo-v2.6-flash-free
$ cd /data/repo && git status --porcelain && git add -A && git commit -q -m "[0.4.230] queue item (x): M27/M28 — the last two case-block exchanges (default<->dns and default<->no_a, closing the exchange family; mutations 388 -> 396/0, 28/28 caught; probe 263/17 and 263/17 recorded before the dict entries existed, control 280/0, both differing-line counts measured 4, untouched middle arm(s) byte-identical at the same offset, queue's 18-component/17-shape figure corrected by measurement) plus the teeth re-read a tenth time, 391/2 when both pins are redirected; PROGRESS entry appended before this commit" && git log --oneline -1 && git status --porcelain; echo "porcelain-empty-rc=$?"
M CHANGELOG.md
 M agent-logs/PROGRESS.md
 M tests/test_ip_drift_cron.sh
 M tools/REGISTRY.md
09bd6e9 [0.4.230] queue item (x): M27/M28 — the last two case-block exchanges (default<->dns and default<->no_a, closing the exchange family; mutations 388 -> 396/0, 28/28 caught; probe 263/17 and 263/17 recorded before the dict entries existed, control 280/0, both differing-line counts measured 4, untouched middle arm(s) byte-identical at the same offset, queue's 18-component/17-shape figure corrected by measurement) plus the teeth re-read a tenth time, 391/2 when both pins are redirected; PROGRESS entry appended before this commit
porcelain-empty-rc=0
> build · mimo-v2.6-flash-free
$ cd /data/repo && echo "--- md5"; md5sum tests/test_ip_drift_cron.sh; git show HEAD:tests/test_ip_drift_cron.sh | md5sum; echo "--- bash -n"; bash -n tests/test_ip_drift_cron.sh && echo OK; echo "--- plain"; bash tests/test_ip_drift_cron.sh 2>&1 | tail -2; echo "--- mutations"; bash tests/test_ip_drift_cron.sh --mutations 2>&1 | tail -2; echo "--- NO_LIVE mutations"; IPDRIFT_NO_LIVE=1 bash tests/test_ip_drift_cron.sh --mutations 2>&1 | tail -2
--- md5
15f0008ccd141c6aac34e55f01353d20  tests/test_ip_drift_cron.sh
15f0008ccd141c6aac34e55f01353d20  -
--- bash -n
OK
--- plain

=== Results: 283 passed, 0 failed ===
--- mutations

=== Results: 396 passed, 0 failed ===
--- NO_LIVE mutations

=== Results: 393 passed, 0 failed ===
> build · mimo-v2.6-flash-free

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

[Tue Oct  6 05:00:11 2026] 127.0.0.1:51564 Accepted
[Tue Oct  6 05:00:11 2026] 127.0.0.1:51564 Closing
[Tue Oct  6 05:00:11 2026] 127.0.0.1:51566 Accepted
[Tue Oct  6 05:00:11 2026] 127.0.0.1:51566 Closing
[Tue Oct  6 05:00:11 2026] 127.0.0.1:51578 Accepted
[Tue Oct  6 05:00:11 2026] 127.0.0.1:51578 Closing
[Tue Oct  6 05:00:11 2026] 127.0.0.1:51588 Accepted
[Tue Oct  6 05:00:11 2026] 127.0.0.1:51588 Closing
[Tue Oct  6 05:00:59 2026] 127.0.0.1:52770 Accepted
[Tue Oct  6 05:00:59 2026] 127.0.0.1:52770 Closing
[Tue Oct  6 05:00:59 2026] 127.0.0.1:52780 Accepted
[Tue Oct  6 05:00:59 2026] 127.0.0.1:52780 Closing
[Tue Oct  6 05:00:59 2026] 127.0.0.1:52784 Accepted
[Tue Oct  6 05:00:59 2026] 127.0.0.1:52784 Closing
[Tue Oct  6 05:00:59 2026] 127.0.0.1:52790 Accepted
[Tue Oct  6 05:00:59 2026] 127.0.0.1:52790 Closing
[Tue Oct  6 05:00:59 2026] 127.0.0.1:52800 Accepted
[Tue Oct  6 05:00:59 2026] 127.0.0.1:52800 Closing
[Tue Oct  6 05:00:59 2026] 127.0.0.1:52802 Accepted
[Tue Oct  6 05:00:59 2026] 127.0.0.1:52802 Closing
[Tue Oct  6 05:01:00 2026] 127.0.0.1:52810 Accepted
[Tue Oct  6 05:01:00 2026] 127.0.0.1:52810 Closing
[Tue Oct  6 05:01:00 2026] 127.0.0.1:52824 Accepted
[Tue Oct  6 05:01:00 2026] 127.0.0.1:52824 Closing
[Tue Oct  6 05:01:00 2026] 127.0.0.1:52832 Accepted
[Tue Oct  6 05:01:00 2026] 127.0.0.1:52832 Closing
[Tue Oct  6 05:01:00 2026] 127.0.0.1:52842 Accepted
[Tue Oct  6 05:01:00 2026] 127.0.0.1:52842 Closing
[Tue Oct  6 05:01:34 2026] 127.0.0.1:44422 Accepted
[Tue Oct  6 05:01:34 2026] 127.0.0.1:44422 Closing
[Tue Oct  6 05:01:34 2026] 127.0.0.1:44434 Accepted
[Tue Oct  6 05:01:35 2026] 127.0.0.1:44434 Closing
[Tue Oct  6 05:01:35 2026] 127.0.0.1:44436 Accepted
[Tue Oct  6 05:01:35 2026] 127.0.0.1:44442 Accepted
[Tue Oct  6 05:01:35 2026] 127.0.0.1:44436 Closing
[Tue Oct  6 05:01:35 2026] 127.0.0.1:44450 Accepted
[Tue Oct  6 05:01:35 2026] 127.0.0.1:44442 Closing
[Tue Oct  6 05:01:35 2026] 127.0.0.1:44458 Accepted
[Tue Oct  6 05:01:35 2026] 127.0.0.1:44450 Closing
[Tue Oct  6 05:01:35 2026] 127.0.0.1:44458 Closing
[Tue Oct  6 05:01:35 2026] 127.0.0.1:44326 Accepted
[Tue Oct  6 05:01:35 2026] 127.0.0.1:44326 Closing
[Tue Oct  6 05:01:35 2026] 127.0.0.1:44332 Accepted
[Tue Oct  6 05:01:35 2026] 127.0.0.1:44332 Closing
[Tue Oct  6 05:04:09 2026] 127.0.0.1:49858 Accepted
[Tue Oct  6 05:04:09 2026] 127.0.0.1:49858 Closed without sending a request; it was probably just an unused speculative preconnection
[Tue Oct  6 05:04:09 2026] 127.0.0.1:49858 Closing
[Tue Oct  6 05:04:10 2026] 127.0.0.1:49860 Accepted
[Tue Oct  6 05:04:10 2026] 127.0.0.1:49860 Closed without sending a request; it was probably just an unused speculative preconnection
[Tue Oct  6 05:04:10 2026] 127.0.0.1:49860 Closing
[Tue Oct  6 05:04:21 2026] 127.0.0.1:44650 Accepted
[Tue Oct  6 05:04:21 2026] 127.0.0.1:44650 Closing
[Tue Oct  6 05:04:45 2026] 127.0.0.1:45690 Accepted
[Tue Oct  6 05:04:45 2026] 127.0.0.1:45690 Closed without sending a request; it was probably just an unused speculative preconnection
[Tue Oct  6 05:04:45 2026] 127.0.0.1:45690 Closing
[Tue Oct  6 05:04:46 2026] 127.0.0.1:45694 Accepted
[Tue Oct  6 05:04:46 2026] 127.0.0.1:45694 Closed without sending a request; it was probably just an unused speculative preconnection
[Tue Oct  6 05:04:46 2026] 127.0.0.1:45694 Closing
[Tue Oct  6 05:09:52 2026] 127.0.0.1:40830 Accepted