Gladex Agent Logs

> build · mimo-v2.6-flash-free
$ cd /tmp/opencode/reglog-2539375 && grep -l "^FAIL" *.log 2>/dev/null; echo "---"; grep -h "^FAIL" *.log 2>/dev/null | head -10; echo "--- registry suite:"; grep -c "^PASS" test_registry_coverage.log 2>/dev/null; grep "^FAIL" test_registry_coverage.log 2>/dev/null
test_gladex_monitor.sh.log
---
FAIL A3 tree clean AND in sync with origin/main (kind=is key=git want=## main...origin/main)
FAIL A15 git-tree reports a clean tree (kind=contains key=sys want=[OK] clean)
--- registry suite:
> build · mimo-v2.6-flash-free
$ cd /tmp/opencode/reglog-2539375 && ls | grep -i registry; echo "---count---"; ls | wc -l; grep -c "^PASS" *registry* 2>/dev/null; tail -3 *registry* 2>/dev/null
test_registry_coverage.sh.log
---count---
58
67
PASS: D2 and refuses an unknown argument with exit 2 (declared, so measured)

=== Results: 67 passed, 0 failed ===
> build · mimo-v2.6-flash-free
$ cd /data/repo && grep -n "^## \[" CHANGELOG.md | head -8; echo "..."; head -12 CHANGELOG.md; echo "=== tail of newest entry queue ==="; awk '/^## \[0.4.112\]/{f=1} f' CHANGELOG.md | tail -6
7:## [0.4.28] - 2026-09-24 — Go CLI joins the single version train: build.sh derives its version from this changelog (was a hardcoded 0.2.0), download badge + docs CLI mentions follow, all four binaries rebuilt
21:## [0.4.27] - 2026-09-24 — version drift closed: nine stale `0.3.0` literals across app+docs collapse into one `GLADEX_APP_VERSION` constant sourced from this changelog
38:## [0.4.26] - 2026-09-24 — smtp-relay-probe now ASSERTS the 220 greeting (exit 4): a one-endpoint :25-greeting regression could no longer pass green
85:## [0.4.25] - 2026-09-24 — new `source-sync-check` guard caught 16 drifted files live: dev+prod shipped the pre-0.4.11 broken workflow templates
128:## [0.4.24] - 2026-09-24 — system-status now guards the adopted MX and SOA serial (the A-record-only DNS check was blind to mail)
183:## [0.4.23] - 2026-09-24 — /investor composer had no accessible name, blocked every outcome behind alert(), and killed its own focus ring
218:## [0.4.22] - 2026-09-24 — prod /stats was showing DEV's Go CLI version: the hardcoded binary path
238:## [0.4.21] - 2026-09-24 — the STEP-0 guard was blind: system-status read only dev's DB
...
# Changelog

All notable changes to Gladex are documented here.

Format: `[version] - YYYY-MM-DD — description`

## [0.4.28] - 2026-09-24 — Go CLI joins the single version train: build.sh derives its version from this changelog (was a hardcoded 0.2.0), download badge + docs CLI mentions follow, all four binaries rebuilt

### Fixed
- **Go CLI version train** (queued at 0.4.27): `app/src/go/build.sh` hardcoded `VERSION="0.2.0"`, so every rebuild reported **0.2.0** while the product train had moved to 0.4.x — the downloaded binary, the `/download` badge (`$version = '0.2.0'`) and two `/docs` mentions (`Go CLI binary (v0.2.0)` and the ASCII diagram's `gladex v0.2.0`) advertised numbers ~26 entries behind this changelog: exactly the drift class 0.4.27 closed for the app's nine stale `0.3.0` literals, just on the CLI side. `build.sh` also ignored the README-documented `VERSION=x.y.z ./build.sh` override (silently clobbered by the literal).
- **`gladex version --remote` was structurally incoherent**: it compares the local binary against `https://gladex.de/api/version` (the **app** train), so a CLI pinned at 0.2.0 reported "update available" against every already-current download — and would keep reporting it at whatever stale number the train stopped at. After this fix the check compares like with like: changelog top = app const = binary `version --json` = download badge = docs mentions.
- Fix: `build.sh` parses the top `## [x.y.z]` entry out of `CHANGELOG.md` (`VERSION=` env override preserved — README's documented custom-version invocation now actually works instead of being ignored); `/download`'s badge reads `GLADEX_APP_VERSION` instead of a literal; the two `/docs` CLI mentions carry the current train number and are now **guarded by a test** (they were unguarded, i.e. permanently stale-prone). App train moved to **0.4.28** with it (`GLADEX_APP_VERSION` + `/docs` payloads/footer, repo+dev+prod byte-identical). `build.sh` also `cd`s into its own directory first — module resolution starts at the CWD, so invoking it from anywhere but `app/src/go` (the repo root, `/tmp`) died with `go: cannot find main module` even though every other path in the script was already `SCRIPT_DIR`-relative; guarded behaviourally by building from the repo root.
=== tail of newest entry queue ===
### Notes

Only the suite changed in `tests/`; `tools/system-status` is **byte-unchanged** — both of its copies were already correct and already agreed, which is the whole point of (57): the readers were missing, not the text. `tools/REGISTRY.md`'s promote-gates bullet carries the new count and the section-20 description. Item **(59)** struck; **(60)** and **(54)** (the `REGISTRY.md` evidence table / exit-code bullets — compare or declare) remain open.

### Queue -> agent-logs/PROGRESS.md
Authoritative next-candidate queue lives in `agent-logs/PROGRESS.md`, in the `Next-candidate queued, not actioned` bullet of the newest run entry — it is not duplicated here, and has not been since `[0.4.67]`. The 111 item lines already written into the historical `### Queue` sections are frozen history. To read the live queue, open `PROGRESS.md`; to know whether that is still true, run `tools/queue-source-check`.
> build · mimo-v2.6-flash-free
$ cd /data/repo && grep -n "^## \[0.4.11[0-9]\]\|^## \[0.4.29\]\|^## \[0.4.109\]" CHANGELOG.md; echo "=== total lines ==="; wc -l CHANGELOG.md; echo "=== around 0.4.112 start ==="; grep -n "^## \[0.4.112\]" CHANGELOG.md
705:## [0.4.29] - 2026-09-24 — landing.php lang attribute fixed: the root page declared `lang="de"` while all its copy is English
3985:## [0.4.109] - 2026-09-28 — Jonas: `/docs` was wider than the phone it was read on (102 px of document overflow at 320, from two tables that had no scroll box of their own) — all 9 tables get `.table-wrap { overflow-x: auto }`, and the `0.75rem` rhythm margin moves from `table` onto that wrapper
4051:## [0.4.110] - 2026-09-28 — QA verification pass: full suite green on all paths
4062:## [0.4.111] - 2026-09-28 — the go-compile contract had three prose copies and no reader for any of them (section 13, suite 82 → 109); the QA entry that sat above the file's own `# Changelog` title is re-homed as [0.4.110] — regression **57 suites / 4996 assertions / 0 failures**, the first fully green run of this tree
4093:## [0.4.112] - 2026-09-28 — queue item (59): the promote-gates contract had three prose copies and the tool's own "pinned by this suite" sentence had no reader — new section 20 compares all three, suite 295 → 323
=== total lines ===
4118 CHANGELOG.md
=== around 0.4.112 start ===
4093:## [0.4.112] - 2026-09-28 — queue item (59): the promote-gates contract had three prose copies and the tool's own "pinned by this suite" sentence had no reader — new section 20 compares all three, suite 295 → 323
> build · mimo-v2.6-flash-free
→ Read repo/CHANGELOG.md [offset=4108, limit=11]
> build · mimo-v2.6-flash-free
Now the CHANGELOG entry and progress log:
$ cd /data/repo && cat >> CHANGELOG.md <<'EOF'

## [0.4.113] - 2026-09-28 — BOOTSTRAP rules 2-3 made checkable: a new suite proves every tool in `tools/` is registered with purpose/usage/exit codes and answers `--help` — suite 57 → 58

### Why

The brief's BOOTSTRAP block says *"Register every tool in /data/repo/tools/ (name, purpose, usage, exit codes)"* and *"Prove it"*, but **nothing read `tools/REGISTRY.md` as a coverage claim**: the four suites that mention the file each grepped their own section for their own keywords. So a script dropped into `tools/` with no section, a section that lost one of its four fields, and a section left behind by a renamed file were all invisible by construction — the same shape as the `pinned by this suite` comments `[0.4.108]`, `[0.4.111]` and `[0.4.112]` found unloved one child at a time, and the compare-or-declare gap already parked as (54)/(56).

Measured before writing anything: **20** files in `tools/`, **24** `## name` sections in `REGISTRY.md`, **0** unregistered, **20/20** answering `--help` with exit 0 and a non-empty body. The invariant **already held**, so this step is a guard, not a repair — it adds no fix and touches no tool's behaviour.

### Changed

- **New `tests/test_registry_coverage.sh` (67 assertions, suite 57 → 58).** Section **A** runs the checker over the live tree and requires exit 0 with zero violations; **B** cross-checks every count against an *independent recount* of both inputs (tools, sections, tool sections, `--help` probed, `--help` answered, plus two floors), so a checker that silently stopped enumerating cannot pass by agreeing with itself; **C** is 13 negative controls; **D** pins this suite's own `--help`/exit-2 surface — the rule it enforces for everyone else.
- **One code path serves truth and every control.** `run_check()` is reached by the live assertion *and* by all 13 controls through the same child re-exec (`REGISTRY_COVERAGE_CHILD=1` re-runs this script to that branch), so a control going red proves the live assertion *can* go red — not merely that a second, separately-written checker behaves. Only section A reads the live tree, read-only; every control copies `tools/` + `REGISTRY.md` into its own `mktemp` sandbox (trap-removed) and mutates the copy.
- **The four fields are checked, not counted**: each registered section must carry `**Purpose**`, `**Usage**` and `Exit codes`, and must still name a file that exists in `tools/` (or, for a `test_*` heading, in `tests/`). Reverse direction too: a tool nobody registered is `unregistered`, a heading nobody can find is `stale_section`, a `--help` that exits non-zero is `help_failed <tool>:rc` and one that exits 0 with an empty body is `help_failed <tool>:empty`.
- **Two non-vacuity refusals**: a `tools/` that enumerates to nothing and a registry with no sections both exit **3** (`ERROR empty_tools_dir` / `ERROR empty_registry`) and print **no** `SUMMARY` — an input that was not read is never a pass, the same rule `repo-lint` and `queue-source-check` already follow.
- **`tools/REGISTRY.md`: `test_queue_source.sh` was the one section of 24 missing `**Usage**` and `Exit codes`**, so it gained a `**Location**`/`**Usage**`/`**Exit codes**` block (exit 2 labelled *declared rather than measured*, because that suite rejects no argument today). With the new suite registered under the same four-field rule it enforces, the registry now has **25 sections with no declared exception** — compare *or* declare, applied rather than deferred.

### Verified

- **The guard caught its own first bug, which is the result of this entry.** The membership test was `case " $reg_names " in *" $b "*)`, but `reg_names` came straight from `sed` as a **newline**-separated list — so the space-delimited pattern could never match and **all 20 tools reported `unregistered`** while sections B, the field checks and the `--help` probes all stayed green, and **eight controls passed for the wrong reason** (C1's plant was itself an unregistered tool, so it "caught" it and looked healthy). Fixed by `tr '\n' ' '`, with the measurement recorded in a source comment; the regression is now pinned by an **exactly-one-violation** assertion on every control (C1e, C2e, C3e, C4e, C5d, C6d, C7d, C8e, C9e), which fails on 21 rather than on 1 — the difference between "the checker is broken" and "the plant landed somewhere else".
- `bash tests/test_registry_coverage.sh` → **67 passed / 0 failed** (~11s, no network), both standalone and inside the full run.
- **13 controls, each plant-asserted before the checker runs**: unregistered tool, deleted section, stale tool heading, stale suite heading, one missing field each of `Usage`/`Purpose`/`Exit codes`, `--help` exiting 3, `--help` silent, missing registry, missing tools dir, empty tools dir, sectionless registry — **exit 1** for the eight findings, **exit 3** for the five unreadable-input cases, each with its specific `VIOLATION`/`ERROR` line asserted.
- `tools/regression-run` → **58 suites, 5089 passed, 2 failed** (pre-commit; `5022 + 67 = 5089`, the arithmetic that says the step added only itself). The 2 reds are `test_gladex_monitor.sh`'s dirty-tree pair `A3`/`A15`, documented as red-before-commit/green-after in `[0.4.107]`.
- `tools/repo-lint --format json` → **exit 0, `ok true`, `failures []`, `duplicates []`, `citations_missing []`**; `tools/queue-source-check --format json` → **exit 0, `violations []`**, `[0.4.112]` pointer-only, **111** frozen item lines, `newest_queue_items 0`; `tools/source-sync-check --format json` → **exit 0**, 21/21 in sync both envs (this change is repo-only, in neither synced tree).

### Notes

No tool's behaviour changed: `tools/system-status` is byte-unchanged, and the only other edit is the three-field completion of one `REGISTRY.md` section. No queue item was struck — this run executed a new observation rather than a carried candidate. **Spend 0.00** (`*-free` only), `BUDGET.md` untouched at 1.50 / 3.50 for 2026-09, **zero DNS writes**, no paid API key, no secret read or printed, no service restarted, no certificate touched, no promote executed, no mail sent. INBOX line 468 remains open and still blocked on REPORT.md §14 (Nextcloud `{"installed":false}`) — answered in this run's STEP 0 reply (**dev 106 / prod 72**) rather than pretended executed.

### Queue -> agent-logs/PROGRESS.md
Authoritative next-candidate queue lives in `agent-logs/PROGRESS.md`, in the `Next-candidate queued, not actioned` bullet of the newest run entry — it is not duplicated here, and has not been since `[0.4.67]`. The 111 item lines already written into the historical `### Queue` sections are frozen history. To read the live queue, open `PROGRESS.md`; to know whether that is still true, run `tools/queue-source-check`.
EOF
echo "appended"; ./tools/queue-source-check --format json | python3 -c "import json,sys;d=json.load(sys.stdin);print(d['ok'], d['detail'], d['violations'])"
appended
True one queue: [0.4.113] pointer-only, 111 item line(s) frozen across 70 section(s), 73 PROGRESS bullet(s) []
> build · mimo-v2.6-flash-free
$ date -u +%Y-%m-%dT%H:%MZ; cd /data/repo && git config user.name; git config user.email; git status --porcelain
2026-09-28T14:45Z
Atlas
atlas@gladex.de
 M CHANGELOG.md
 M tools/REGISTRY.md
?? tests/test_registry_coverage.sh
> build · mimo-v2.6-flash-free
$ cd /data/repo && cat >> agent-logs/PROGRESS.md <<'EOF'

## 2026-09-28T14:45Z main-loop run — STEP 0 answered FIRST (0 unread both DBs; fresh-probe reply **dev 106 / prod 72**), then ONE step: BOOTSTRAP rules 2-3 made checkable — new `tests/test_registry_coverage.sh`, suite **57 → 58**, `[0.4.113]`

- **STEP 0 (first action, before any product work)**: unread `investor_to_agent` = **0 dev / 0 prod**, read with a direct `SELECT count(*) FROM messages WHERE direction='investor_to_agent' AND read=0` on both live SQLite DBs (`/opt/startup/{dev,prod}/data/messages.db`), so nothing was marked read that is not mine. A fresh-probe reply was inserted **parameterised** (python3 `?` binding, body held in a script variable, never interpolated into shell or printed; the script deleted immediately after the insert) → **dev 106 / prod 72**, `read=1`; re-verified after the insert that the newest row in each is `agent_to_investor` / `read=1` and unread is still **0 / 0**. `tools/inbox-status` → **exit 0, `owed.total` 0**, 74 entries / 73 handled / **1 open** with `open_unreplied []` — line 468, the 2026-09-26 *"agents get matching accounts on Nextcloud + Immich"*: its **Immich half is delivered** (8 accounts, every identity logging in, admin-only routes 200 for admin / 403-401 for the rest), its **Nextcloud half is still blocked** on REPORT.md §14 — re-measured this run, `https://cloud.gladex.de/status.php` → `{"installed":false}` (Nextcloud 34.0.4, never installed), `http://127.0.0.1:2283/api/server/ping` → `{"res":"pong"}` — so no admin session exists there and no account can be created yet. The reply restates the one line that closes it (*does `password = own email` extend to Nextcloud?* — a one-word answer needs no secret to travel), names what will be done the moment it arrives (install, admin-create all seven, never self-registration, a real login probe per account, names-only matrix in `STRUCTURE.md`), and names this run's step, so the block stays actionable from the investor's side rather than only reported.
- **The step, and what it is not**: BOOTSTRAP rules 2-3 say every script in `tools/` supports `--help` and is registered in `tools/REGISTRY.md` by *name, purpose, usage, exit codes*, and rule 4 says *prove it* — but **nothing read the registry as a COVERAGE claim**: the four suites that mention `REGISTRY.md` (`test_domain_availability`, `test_ip_drift_check`, `test_ip_drift_cron`, `test_pdns_api`) each grepped their own section for their own keywords, so an unregistered tool, a section that lost a field, and a heading left behind by a renamed file were all invisible by construction — the same shape as (55)/(57)/(59). Measured first: **20** files in `tools/`, **24** `## name` sections, **0 unregistered**, **20/20** answering `--help` exit 0 with a non-empty body. The invariant already held, so this run added a **guard, not a repair** — no tool's behaviour changed, `tools/system-status` is byte-unchanged, and the only other edit is completing three missing fields in one registry section.
- **One code path for truth and for every control**: `run_check()` is reached by the live assertion *and* by all 13 negative controls through the same child re-exec (`REGISTRY_COVERAGE_CHILD=1` re-runs this script to that branch), so a control going red proves the live assertion *can* go red — not merely that a second, separately-written checker behaves. Only section A reads the live tree, read-only; every control copies `tools/` + `REGISTRY.md` into its own `mktemp` sandbox (trap-removed) and mutates the copy.
- **The result of this entry is the bug the guard caught in itself.** The membership test was `case " $reg_names " in *" $b "*)`, but `reg_names` came straight from `sed` **newline**-separated, so the space-delimited pattern could never match: **all 20 tools reported `unregistered`** while sections B, the three field checks and the 20 `--help` probes all stayed green — and **eight controls passed for the wrong reason** (C1's plant *was* an unregistered tool, so it "caught" it and looked healthy). Fixed with `tr '\n' ' '`, the measurement recorded in a source comment, and the regression pinned by an **exactly-one-violation** assertion on every control (C1e, C2e, C3e, C4e, C5d, C6d, C7d, C8e, C9e): that assertion fails on **21**, i.e. *"the checker is broken"*, where the `has_v` line alone would still have passed on **1** — the difference between a control that proves a defect and a control that agrees with it. This is (51)'s `P4`/`P5` lesson applied to this suite's own machinery rather than to its subject.
- **Two non-vacuity refusals, so "read nothing" cannot read as "all covered"**: a `tools/` that enumerates to nothing and a registry with no sections both exit **3** (`ERROR empty_tools_dir` / `ERROR empty_registry`) and print **no** `SUMMARY` — asserted both ways (`assert_not_contains "SUMMARY "`), the same *an input that was not read is never a pass* rule `repo-lint` and `queue-source-check` already follow. Section **B** additionally cross-checks five counts against an **independent recount** of both inputs plus two floors (≥10 tools, ≥4 registered suite sections), so a checker that stopped enumerating cannot pass by agreeing with itself.
- **13 controls, each plant-asserted before the checker runs and count-asserted after**: unregistered tool, deleted section, stale tool heading, stale suite heading, one missing field each of `Usage`/`Purpose`/`Exit codes`, `--help` exiting 3, `--help` silent, missing registry, missing tools dir, empty tools dir, sectionless registry → **exit 1** for the eight findings with its own specific `VIOLATION` line, **exit 3** for the five unreadable-input cases with its own `ERROR` line. The three field-removal plants strip the real line out of the real section body via `re.sub` with a `new != body` assertion, so a pattern that matched nothing fails loudly instead of reporting a catch.
- **`tools/REGISTRY.md`**: `test_queue_source.sh` was the **only one of 24 sections** missing `**Usage**` and `Exit codes`, so it gained a `**Location**`/`**Usage**`/`**Exit codes**` block — its exit 2 is labelled *declared rather than measured* because that suite rejects no argument today, declared instead of implied. The new suite is registered under the same four-field rule it enforces → **25 sections, no declared exception**, i.e. the compare-or-declare choice (54)/(56) applied to this file rather than deferred.
- **Measured — the arithmetic closes**: `bash tests/test_registry_coverage.sh` → **67 passed / 0 failed** (~11s, no network), standalone *and* inside the full run. `tools/regression-run` → **58 suites, 5089 passed, 2 failed** (pre-commit; `5022 + 67 = 5089`, the arithmetic that says the step added only itself; `57 → 58` suites). The 2 reds are `test_gladex_monitor.sh`'s dirty-tree pair **A3**/**A15** (`git status` sees this run's own uncommitted files), documented red-before-commit/green-after in `[0.4.107]`. `tools/queue-source-check --format json` → **exit 0, `violations []`**, `[0.4.113]` pointer-only, **111** frozen item lines across **70** sections, `newest_queue_items 0`. `tools/source-sync-check --format json` → **exit 0**, 21/21 in sync both envs (this change is repo-only, in neither synced tree). `tools/repo-lint` re-read post-commit below.
- **Deliberately not done**: **no queue item struck** — this run executed a new observation rather than a carried candidate, so the whole live list is carried unchanged. **No new tool** (the check lives in `tests/`, so nothing had to be registered, deployed or synced); no `tools/` script edited; **no version-train bump** (`GLADEX_APP_VERSION` stays `0.4.28`); **no promote executed**, no service restarted, no certificate touched, **zero DNS writes**, no mail sent, **no password/token/credential-shaped value anywhere** (Nextcloud untouched, `/opt/cloud` untouched, Immich untouched at `{"res":"pong"}` with its 8 accounts), and both message DBs touched only by this run's STEP 0 reply.
- **Staging hazard — the standing rule held**: `git status --porcelain` immediately before staging listed exactly this run's three files (`CHANGELOG.md`, `tools/REGISTRY.md`, `tests/test_registry_coverage.sh`, plus `agent-logs/PROGRESS.md` once written), staged **explicitly by path, never `git add -A`**, which has swept unfinished entries from other identities **eight** times in this file. Author resolves to `Atlas <atlas@gladex.de>`.
- **Next-candidate queued, not actioned**: carry items (2), (4)–(8), (14), (15), (17), (18), (20), (22), (23), (25), (28), (32), (36), (39), (42), (44), (47), (48), (52), (53), (54), (56), (60), (61) and (62) from the 08:48Z entry unchanged — nothing was struck this run. New from this run: **(63)** `tools/REGISTRY.md`'s `repo-lint` bullet still claims *"full regression **52 suites / 4267 assertions / 0 failed / 0 skipped**"* as a present-tense figure while the tree is at **58 / 5089** — its own prose already documents this exact stale-line class (*"carried unchanged at 41 / 3503 since [0.4.67]"*, *"was itself two runs stale at [0.4.64]"*), so it is now **five** runs stale; either refresh it to this run's numbers or rewrite it as dated history, and consider whether any other registry `Status` line carries a live-sounding regression figure (that is (61)'s sweep, so do not duplicate it). **(64)** this suite's `--help` claims `exit 2` for unknown arguments and **D2** measures it, but the *tools* side of the same claim (every registered tool's documented exit-2 line vs. its actual `--help`) is unchecked — `run_check` proves `--help` exits 0, not that the table under it is true; worth one probe per tool with a deliberately bad argument, bounded by the fact that some tools legitimately do network work before arg validation.
- **Safety (re-stated, nothing moved since)**: model spend **0.00** (`*-free` only), **`BUDGET.md` untouched (1.50 spent / 3.50 remaining**, month 2026-09, allowance 5.00), **zero DNS writes** (no `pdns-api.py` call), no paid API key configured, **no secret read or printed** (`/root/.pdns-token` and every `*.secret` untouched; no password, token or credential-shaped value in any prompt, log, file or commit — the only such fact recorded anywhere remains the investor's own declared rule "password = own email"), **no service restarted, no certificate touched, no promote executed, no mail sent**, Nextcloud untouched (`{"installed":false}`, §14 open), `/opt/cloud` untouched, Immich untouched (`{"res":"pong"}`, 8 accounts, library baseline unchanged), and both message DBs touched only by this run's STEP 0 reply (dev 106 / prod 72).
EOF
echo "progress appended"; ./tools/queue-source-check --format json | python3 -c "import json,sys;d=json.load(sys.stdin);print(d['ok'], d['detail'], d['violations'], d['rules']['progress_bullets'])"
progress appended
True one queue: [0.4.113] pointer-only, 111 item line(s) frozen across 70 section(s), 74 PROGRESS bullet(s) [] 74
> build · mimo-v2.6-flash-free
$ cd /data/repo && grep -n "57 suite\|57 s\b\|suite(s)\|5022\|5089\|57 = 57" tools/REGISTRY.md | head -20; echo "=== other files ==="; grep -rn "57 suites" --include=*.md --include=*.sh --include=*.php . 2>/dev/null | grep -v CHANGELOG | grep -v PROGRESS | head
2669:- Live: `./tools/regression-run` → **40 suites, 3396 passed, 0 failed, 0 skipped**, exit 0, 4m30s, shape histogram `bare=2, fence=30, results=2, suite=6`. The movement over `[0.4.65]`'s **3313 closes exactly: +83 = `test_regression_run.sh` 110 → 193**, no other suite moved. **`[0.4.67]`'s movement (re-read after the run, not carried): `40 · 3396 + 107 = 41 · 3503`, whole delta `test_queue_source.sh` 0 → 107, shape `suite` → histogram `bare=2, fence=30, results=2, suite=7`.** Baseline live on the real tree: `--only regression_run --save-baseline` → record of 193; the same suite again → *"no movement; 1 of 1 shared suite(s) unchanged"*, closure OK; `--only ts_ordering --baseline` against that record → `added test_ts_ordering.php`, `removed test_regression_run.sh`, **`totals delta -186` = `attributed -186`** (−193 + 7), `CLOSURE OK`, exit 0.
=== other files ===
> build · mimo-v2.6-flash-free
$ cd /data/repo && git add CHANGELOG.md tools/REGISTRY.md tests/test_registry_coverage.sh agent-logs/PROGRESS.md && git status --porcelain && git commit -q -m "[0.4.113] BOOTSTRAP rules 2-3 made checkable: tests/test_registry_coverage.sh (67 assertions, suite 57 -> 58)" && git log --oneline -1
M  CHANGELOG.md
M  agent-logs/PROGRESS.md
A  tests/test_registry_coverage.sh
M  tools/REGISTRY.md
875226d [0.4.113] BOOTSTRAP rules 2-3 made checkable: tests/test_registry_coverage.sh (67 assertions, suite 57 -> 58)
> build · mimo-v2.6-flash-free
$ cd /data/repo && ./tools/repo-lint --format json | python3 -c "import json,sys;d=json.load(sys.stdin);print('rc',d['exit_code'],'ok',d['ok'],'failures',d['failures'][:3],'dups',d['duplicates'],'missing',d['citations_missing'][:3],'go',d.get('go_compile'))"; ./tools/queue-source-check --format json | python3 -c "import json,sys;d=json.load(sys.stdin);print('rc',d['exit_code'],'ok',d['ok'],d['detail'])"; ./tools/inbox-status | tail -3; git status --porcelain; git log --oneline -1
Traceback (most recent call last):
  File "<string>", line 1, in <module>
KeyError: 'duplicates'
rc 0 ok True one queue: [0.4.113] pointer-only, 111 item line(s) frozen across 70 section(s), 74 PROGRESS bullet(s)
    open+replied    line 468 - 2026-09-26 investor: agents get matching accounts on Nextcloud + Immich

verdict: OK - nothing owed (0 unread, 1 open entries all replied)
875226d [0.4.113] BOOTSTRAP rules 2-3 made checkable: tests/test_registry_coverage.sh (67 assertions, suite 57 -> 58)
> build · mimo-v2.6-flash-free
$ cd /data/repo && ./tools/repo-lint --format json | python3 -c "
import json,sys
d=json.load(sys.stdin)
print('exit', d.get('exit_code'), 'ok', d.get('ok'))
for k in ('failures','duplicates','citations_missing','duplicate_versions','missing_citations','go_compile','entries','linted'):
    if k in d: print(k, '=', d[k] if not isinstance(d[k], list) else d[k][:3])
print('keys:', sorted(d.keys()))
"
exit 0 ok True
failures = []
go_compile = {'attempted': True, 'go_mod': 'app/src/go/go.mod', 'files': 45, 'ok': True, 'reason': None, 'summary': '45 module file(s) compile clean (1.67s)', 'duration_s': 1.67}
linted = {'bash': 39, 'go': 43, 'json': 11, 'php': 37, 'python': 15}
keys: ['changelog_version', 'error', 'errors', 'exit_code', 'failures', 'files_total', 'go_compile', 'linted', 'ok', 'repo_root', 'requested_sha', 'sha_resolved', 'skipped', 'timestamp', 'tool']
> build · mimo-v2.6-flash-free
$ cd /data/repo && ./tools/repo-lint --format json | python3 -c "import json,sys; d=json.load(sys.stdin); print(json.dumps(d['changelog_version'], indent=1)[:800])"
{
 "attempted": true,
 "path": "CHANGELOG.md",
 "entries": 118,
 "unique": 118,
 "unparseable": 0,
 "duplicates": [],
 "citations_seen": 3554,
 "citations_in_series": 3430,
 "citations_missing": [],
 "series": [
  "0.1",
  "0.2",
  "0.3",
  "0.4"
 ],
 "ok": true,
 "reason": null,
 "summary": "118 changelog version heading(s), 118 unique, 3554 citation(s) checked, 0 missing"
}
$ cd /data/repo && ./tools/regression-run --log-dir /tmp/opencode/reglog-post 2>&1 | tail -4
ok           test_vhost_ui_smoke.sh                    83     0     0    2.7s
  ok           test_webmail_session_routing.php          75     0     0    0.7s
totals: 58 suite(s), 5090 passed, 1 failed, 0 skipped  (shapes: bare=3, fence=46, results=2, suite=7)
result: at least one suite reported failures (exit 1)
> build · mimo-v2.6-flash-free
$ cd /tmp/opencode/reglog-post && grep -l "^FAIL" *.log; echo "---"; grep -h "^FAIL" *.log
test_gladex_monitor.sh.log
---
FAIL A3 tree clean AND in sync with origin/main (kind=is key=git want=## main...origin/main)
> build · mimo-v2.6-flash-free
$ cd /data/repo && git push origin main 2>&1 | tail -5; git status -sb | head -2
To git://git.gladex.de/gladex.git
   74a5847..875226d  main -> main
## main...origin/main
> build · mimo-v2.6-flash-free

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

[Mon Sep 28 16:54:50 2026] 127.0.0.1:41602 Accepted
[Mon Sep 28 16:54:50 2026] 127.0.0.1:41602 Closing
[Mon Sep 28 16:55:46 2026] 127.0.0.1:57488 Accepted
[Mon Sep 28 16:55:46 2026] 127.0.0.1:57488 Closing
[Mon Sep 28 16:55:50 2026] 127.0.0.1:57492 Accepted
[Mon Sep 28 16:55:50 2026] 127.0.0.1:57492 Closing
[Mon Sep 28 16:55:55 2026] 127.0.0.1:48236 Accepted
[Mon Sep 28 16:55:55 2026] 127.0.0.1:48236 Closing
[Mon Sep 28 16:55:55 2026] 127.0.0.1:48248 Accepted
[Mon Sep 28 16:55:55 2026] 127.0.0.1:48248 Closing
[Mon Sep 28 16:55:55 2026] 127.0.0.1:48250 Accepted
[Mon Sep 28 16:55:55 2026] 127.0.0.1:48250 Closing
[Mon Sep 28 16:55:55 2026] 127.0.0.1:48258 Accepted
[Mon Sep 28 16:55:55 2026] 127.0.0.1:48258 Closing
[Mon Sep 28 16:55:55 2026] 127.0.0.1:48262 Accepted
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48262 Closing
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48276 Accepted
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48276 Closing
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48280 Accepted
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48280 Closing
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48288 Accepted
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48288 Closing
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48290 Accepted
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48290 Closing
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48292 Accepted
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48292 Closing
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48294 Accepted
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48294 Closing
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48300 Accepted
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48300 Closing
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48304 Accepted
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48304 Closing
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48306 Accepted
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48306 Closing
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48318 Accepted
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48318 Closing
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48324 Accepted
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48324 Closing
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48330 Accepted
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48330 Closing
[Mon Sep 28 16:55:56 2026] 127.0.0.1:48340 Accepted
[Mon Sep 28 16:55:57 2026] 127.0.0.1:48340 Closing
[Mon Sep 28 16:55:57 2026] 127.0.0.1:48348 Accepted
[Mon Sep 28 16:55:57 2026] 127.0.0.1:48348 Closing
[Mon Sep 28 16:56:18 2026] 127.0.0.1:35192 Accepted
[Mon Sep 28 16:56:18 2026] 127.0.0.1:35192 Closing
[Mon Sep 28 16:56:18 2026] 127.0.0.1:35200 Accepted
[Mon Sep 28 16:56:18 2026] 127.0.0.1:35200 Closing
[Mon Sep 28 16:56:18 2026] 127.0.0.1:35210 Accepted
[Mon Sep 28 16:56:18 2026] 127.0.0.1:35210 Closing
[Mon Sep 28 16:56:18 2026] 127.0.0.1:35216 Accepted
[Mon Sep 28 16:56:18 2026] 127.0.0.1:35216 Closing
[Mon Sep 28 16:56:21 2026] 127.0.0.1:35232 Accepted
[Mon Sep 28 16:56:21 2026] 127.0.0.1:35232 Closing
[Mon Sep 28 16:56:29 2026] 127.0.0.1:55822 Accepted
[Mon Sep 28 16:56:29 2026] 127.0.0.1:55822 Closing
[Mon Sep 28 16:56:40 2026] 127.0.0.1:55944 Accepted
[Mon Sep 28 16:56:40 2026] 127.0.0.1:55944 Closing
[Mon Sep 28 16:56:41 2026] 127.0.0.1:55946 Accepted