gflow-cli 0.82.0__tar.gz → 0.82.1__tar.gz

{gflow_cli-0.82.0 → gflow_cli-0.82.1}/.codex-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "gflow",
-  "version": "0.82.0",
+  "version": "0.82.1",
   "description": "Reusable development and operations workflows for gflow-cli",
   "author": {
     "name": "Flavio Oliva",

{gflow_cli-0.82.0 → gflow_cli-0.82.1}/CHANGELOG.md

@@ -7,6 +7,42 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [0.82.1] — 2026-10-02
+
+### Fixed
+
+- **A video's Flow workflow id is recorded in the catalog (#898).** It was written as
+  `NULL` for every clip. As a result, the MCP task result's `flow_workflow_id` was always
+  `null` for a video, and a clip could not be looked up by its workflow id. On
+  flow.google.com the id now comes from the generation record (slot 0, the id Flow polls
+  the clip by); on labs, from the generate reply's `media[0].workflowId`, which is not the
+  operation name already stored as `flow_operation_id`.
+- **Recording a started video twice no longer crashes (#898).** A second start for a
+  media id already in the catalog minted a fresh row id and hit
+  `UNIQUE(profile_name, flow_media_id)` (`DataIntegrityError`). A start for a media id the
+  catalog already holds is now a no-op, so it cannot reset a completed clip to `pending`.
+- **A busy catalog no longer turns a successful generation into a failed one (#900).**
+  When another gflow process held the local database's write lock past the 5 s
+  `busy_timeout`, recording a finished clip raised a raw `sqlite3.OperationalError`. That
+  missed every handler that downgrades a recording failure to a warning, so `gflow video`
+  exited 1 ("unexpected error") for a clip that existed and was paid for, and tried to
+  record it as failed. Catalog transactions now raise `DataStoreError`, so `gflow image`
+  and `gflow video` succeed and log a `data.persistence_failed_after_success` warning
+  naming the media id. Outside a successful generation the same failure exits 16
+  instead of 1.
+- **A reCAPTCHA mint failure is a typed error with Problem Details (#915).** It was a bare
+  `RuntimeError`: `gflow image upscale` and `gflow video extend` (and image generation on
+  the experimental HTTP transports) exited 1 with an "unexpected" message and no
+  remediation, an MCP/worker caller got a hashed "Unknown Error", and on an experimental
+  transport one failure ended a multi-prompt run past `--continue-on-error`. It now has its
+  own `type` (`…/errors/recaptcha-mint`) and a `retryable` flag set from a live measurement:
+  a mint that lost a race with a navigation, or ran on a Flow page before Flow injected its
+  reCAPTCHA script, is retryable (the settled page mints); a mint on a page that is not a
+  web page (`about:blank`) is not (it fails the same way every time); an empty token, never
+  observed, keeps the class default (not retryable). It still exits 1 —
+  branch on the `type`. The site-key message no longer blames "the Flow editor page" or "the
+  script tag layout", which was false on the `about:blank` page where it was seen (#891).
+
 ## [0.82.0] — 2026-10-01
 
 ### Added
@@ -5784,7 +5820,8 @@ shell-script template that branches on these codes.
 
 First skeleton. Not functional end-to-end yet.
 
-[Unreleased]: https://github.com/ffroliva/gflow-cli/compare/v0.82.0...HEAD
+[Unreleased]: https://github.com/ffroliva/gflow-cli/compare/v0.82.1...HEAD
+[0.82.1]: https://github.com/ffroliva/gflow-cli/compare/v0.82.0...v0.82.1
 [0.82.0]: https://github.com/ffroliva/gflow-cli/compare/v0.81.0...v0.82.0
 [0.81.0]: https://github.com/ffroliva/gflow-cli/compare/v0.80.0...v0.81.0
 [0.80.0]: https://github.com/ffroliva/gflow-cli/compare/v0.79.1...v0.80.0

{gflow_cli-0.82.0 → gflow_cli-0.82.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: gflow-cli
-Version: 0.82.0
+Version: 0.82.1
 Summary: CLI and MCP server for Google Flow — drive Veo text-to-video, image-to-video and Imagen image generation from the terminal.
 Project-URL: Homepage, https://github.com/ffroliva/gflow-cli
 Project-URL: Documentation, https://ffroliva.github.io/gflow-cli/

{gflow_cli-0.82.0 → gflow_cli-0.82.1}/docker/Dockerfile

@@ -45,7 +45,7 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
 # Deliberately NOT pinned: google-chrome-stable. Chrome's apt repo keeps only the current
 # build, so a pinned version stops resolving within weeks. The trade-off is that this
 # layer caches indefinitely and a rebuild is what picks up a newer Chrome.
-ARG GFLOW_VERSION=0.82.0
+ARG GFLOW_VERSION=0.82.1
 RUN pip install --no-cache-dir "gflow-cli==${GFLOW_VERSION}"
 
 VOLUME ["/data"]

{gflow_cli-0.82.0 → gflow_cli-0.82.1}/docs/DATA_LAYER.md

@@ -239,6 +239,7 @@ The data layer follows a "fail fast before billing, warn after success" contract
 | DB has a NEWER schema than installed gflow-cli | Raise `DataMigrationError` BEFORE Flow call | 16 |
 | Recorder write fails AFTER successful paid generation | Emit `data.persistence_failed_after_success` structlog event with `flow_media_id` + `local_path` when available; print yellow warning; **return 0** | 0 |
 | Batch row N fails to persist | Same warning; later rows still recorded | 0 (or 1 if other failure modes apply) |
+| A write cannot take the lock within `busy_timeout` (another process is writing) | `DataStore.transaction()` raises `DataStoreError` (#900): after a successful generation, the warning row above; anywhere else, the error | 0 after success, else 16 |
 
 **Why "warn and continue" after success?** If your `gflow video t2v` succeeds and the file lands on disk, exiting non-zero solely because the local catalog could not be updated would teach scripts to retry the paid generation. The catalog can be reconciled later by a future `gflow data repair`; the file cannot be un-billed.
 
@@ -521,7 +522,7 @@ Every SQLite connection opened by `DataStore` enables:
 |---|---|---|
 | `foreign_keys` | `ON` | The schema declares FK constraints; SQLite ignores them by default |
 | `journal_mode` | `WAL` | Allows parallel CLI processes to read while one writes |
-| `busy_timeout` | `5000` (ms) | Two CLI processes upserting concurrently retry briefly instead of failing immediately with `database is locked` |
+| `busy_timeout` | `5000` (ms) | Two CLI processes upserting concurrently retry briefly instead of failing immediately with `database is locked`; past 5 s the write raises `DataStoreError` (#900) |
 
 Writes are explicitly grouped with `BEGIN IMMEDIATE` via `DataStore.transaction(immediate=True)` to avoid deferred-writer deadlocks under WAL. Connections use `isolation_level=None` so the transaction context manager controls boundaries explicitly.
 

{gflow_cli-0.82.0 → gflow_cli-0.82.1}/docs/INDEX.md

@@ -171,7 +171,7 @@ Slash commands for Claude Code, stored in `.claude/commands/gflow/`. All prefixe
 **"A command failed — where is the incident bundle and what's in it?"** → [DEBUGGING § Automatic incident bundles](DEBUGGING.md#automatic-incident-bundles) (layout, triggers, review-before-sharing); privacy boundaries: [SECURITY § Automatic incident bundles](SECURITY.md#automatic-incident-bundles-gflow_cli_incident_capture-default-on); disable via [CONFIGURATION § GFLOW_CLI_INCIDENT_CAPTURE](CONFIGURATION.md#gflow_cli_incident_capture)
 **"Flow's UI broke a selector — how do I diagnose it?"** → [DEBUGGING § Inspecting Flow's live UI](DEBUGGING.md#inspecting-flows-live-ui)
 **"What does each `ui_automation.*` log event mean?"** → [DEBUGGING § Listener & HTTP-layer debugging](DEBUGGING.md#listener--http-layer-debugging)
-**What was actually live-verified for the latest release?** → latest: [LIVE_VERIFICATION_v0.82.0](LIVE_VERIFICATION_v0.82.0.md) (**2026-10-01.** `"ref": "batch:N"` and local-file refs on `gflow run` verified by a BDD e2e on flow.google.com: in place, no re-upload, lineage recorded; 1 veo-lite clip for the #926 e2e) · previous: [LIVE_VERIFICATION_v0.81.0](LIVE_VERIFICATION_v0.81.0.md) (**2026-09-30, 3 veo-lite clips.** The off-screen generation browser (#923) measured unchanged for image and video, window placement read back from Chrome, focus-steal measured and documented; #919 verified at merge) · previous: [LIVE_VERIFICATION_v0.80.0](LIVE_VERIFICATION_v0.80.0.md) (**2026-09-29, 0 Veo credits.** `--resolution` 360p/720p proven by driving the real settings pane on `omni_flash` and refusing the credit-spending submit, on the CLI transport **and** the MCP twin; `nano2-lite` e2e 2 passed; #891 measured before/after — `gflow run` `[nano2, imagen4]` went from exit 1 `RecaptchaError` with a lost results table to exit 36 naming the model, plus 3 e2e (fresh, warm, MCP) with a spy asserting zero client mints; #896 `e2e_data` 2 passed, its failure path offline-only. Named blockers: rendered resolution (needs a billed clip), the `/about` re-check arm, labs-served accounts.) · previous: [LIVE_VERIFICATION_v0.79.1](LIVE_VERIFICATION_v0.79.1.md) (**the signed-media download survives a dropped connection, verified 2026-09-22 at $0.** All three download sites route through one new shared helper, and that helper was driven through a real Flow clip route and a real signed CDN transfer by `pytest -m e2e_data` — 2 passed in 20.74 s, `ftyp` magic, byte length equal to the size Flow reports, and **no `url=` on any log event**. The retry itself could not wait for an organic reset — 20/20 clean here against the reporter's 2-in-6 — so it was **measured against a real TCP RST with an A/B control**: with `max_retries` the transfer recovers on a second connection, with it disabled the identical fault kills the call, and three attempts at a 5 s timeout finish in 1.57 s rather than ~15 s, proving the backoff shares one budget. **Two sites are NOT live-verified and name their blockers:** the composer call site needs a billed generation and `gflow credits` answers 401 on this account (#795), so no credits were spent; and the labs download cannot be reached at all from here (HTTP 308, 6/6 profiles). Offline: 21 new tests, full sweep 2143 passed.) · earlier: [LIVE_VERIFICATION_v0.79.0](LIVE_VERIFICATION_v0.79.0.md) (**the migrated host stops echoing a model it never observed, verified 2026-09-18 at $0.** `model_name_type` is `null` on `flow.google.com` in a real `image t2i --json`, with every sibling field populated — which is what makes the null a result rather than an empty envelope; the first read of it was flat rather than into `images[]` and returned `None` for everything, and that instrument error was discarded rather than reported. Both surfaces ran: `test_e2e_t2i_runs_on_a_moved_account` and `test_e2e_mcp_i2i_runs_on_the_migrated_host`, 2 passed in 86 s. **`gflow data download` was tested on an image and failed** — 45 s, then a message blaming the project, the trash and your prompt; recovery is video-only and the release notes now say so (#877). **#803 is NOT live-verified and names its blocker:** reaching those raise sites needs a non-JSON or token-less aisandbox reply, and the labs route refuses first with a 404 — itself a misleading error, filed as #875. Offline: 4 new unit tests, 1775 passed.) · previous: [LIVE_VERIFICATION_v0.78.0](LIVE_VERIFICATION_v0.78.0.md) (**generating with no project on `flow.google.com` — and video from a start frame — verified 2026-09-17.** The release exists because labs.google's `project.createProject` is retired — 404 *"Flow RPCs have been deprecated and disabled"* with a labs token, 401 without, 0 successes in 12 across four profiles — and gflow reported the 401 as an expired login. Creation and rename are driven on flow.google.com and read back from Flow's own `jHPbke` / `o8DA4` replies; the CLI and the MCP twin are each exercised separately, and both video paths reach a real submit that is aborted before Flow can bill it. Two failures are recorded and re-tested rather than explained away — one MCP image reply that did not reproduce under a four-run A/B, and a Chrome launch starved by a concurrent coverage run that passed in 36 s alone. The 3:4 image aspect is re-measured on the live radiogroup, five radios where 2026-09-08 saw four. Three items are NOT verified and each names its blocker: no labs-served account, no entitled `/about` account, no non-`en` locale. **#860 is the one thing here that cost credits, deliberately:** the only proof that a start frame binds is a real generation, so ten i2v clips were generated and downloaded on a live migrated account — `migrated.frame_uploaded` → `frame_bound` → `submit_observed` → `download`, the middle event being the one that could not fire before this release. **#859 is recorded as UNOBSERVED, not verified** — its dismissal ran on every one of those runs and no promo modal ever appeared, so the $0 route-intercepted e2e is what covers it. `gflow docs` (#861) is verified from the built wheel, which is its whole contract: 127 topics, `docs/FILE.md:LINE` search hits, no network and no checkout.) · previous: [LIVE_VERIFICATION_v0.77.1](LIVE_VERIFICATION_v0.77.1.md) (**a four-fix patch release, verified 2026-09-17 at ZERO cost.** The headline is #846: the published package could not run a single interactive command on a clean Windows install, because `ConsoleRenderer(colors=True)` needs `colorama` and nothing declared it. Verified as a controlled A/B on a fresh Python 3.13 venv — `SystemError` before, exit 33 after — plus an arm with stderr piped that shows why the AUTO/JSON split hid this from every CI job, and the MCP twin run separately. #848's import probe is exercised against a **deliberately broken venv** and an intact one, including the contract that a probe which cannot run reports nothing rather than reporting breakage. #852 is re-measured on the release head: 40 000 base64-shaped chars fall from 12.07 s to 0.020 ms with identical output on Workspace and custom domains. **#849 is a NAMED BLOCKER, not an omission** — its end-to-end behaviour needs a fresh interactive Google sign-in, which no agent can perform; its live A/B is recorded on PR #851 and its e2e file re-ran 4/4 here. **Seven live e2e failures are recorded rather than hidden**, and controlled: the release changes exactly four source files and none of them is a failing test's module.) · previous: [LIVE_VERIFICATION_v0.77.0](LIVE_VERIFICATION_v0.77.0.md) (**the migrated-host login lockout (#791), verified as a controlled A/B on a real Google **Workspace** account at ZERO cost — same profile, same cookies untouched between arms, so the only variable is the code: exit 8 with the Chrome marker rolled back and `channel_for_profile` collapsed to `None`, then exit 0 with the marker intact and the account recorded. The refusal path is measured too — of 10 local profiles, 5 reproduce #791 and exactly one carries both required cookies: 4 refusals, 1 upgrade, 0 false positives. v0.76.0 recorded this fix as **withdrawn**; the approach that shipped is a different one, server-attested on the landing URL rather than on client-rendered DOM. An earlier revision gated auth on an `@gmail.com`-only pattern that silently declined every custom domain — the account used here is one, which is why this is its verification. **#796 is reproduced and prevented**, not merely reported: it was filed as *'not yet reproduced end-to-end'*. **Two rows are named blockers, not omissions:** the MCP Registry publish (#841) cannot be rehearsed before a real tag push, so this release is its first exercise; and the migrated-host generation e2e cannot run here at all — every account bounces to `/about`, A/B-controlled against `develop` so it is not mistaken for a regression.) · previous: [LIVE_VERIFICATION_v0.76.0](LIVE_VERIFICATION_v0.76.0.md) (**a diagnostics-and-distribution release, verified 2026-09-16 at ZERO cost — no Veo credits, no image generation, no Flow submit.** The decisive arm is #833's **negative** control, run first-hand on two real accounts rather than relayed: `flow-pinhole-unavailable-screen` is **0** on both `ffroliva` and `denon82` while `aisandbox-root` is **1**, so the guard does not fire on accounts that HAVE Flow access and the zero is a real absence rather than a page that never booted — a guard with only a positive arm is indistinguishable from one that is permanently on. Both accounts settled at **`/about`**, which is the state that broke this release's *withdrawn* #791 oracle, and the exit-39 signal reads a component that only exists on the unavailable screen, so `/about` cannot mimic it. mcpservers.org is verified with a **404 negative control** because the listing 403s to `curl` and a naive check would have called the link dead. **#791 is NOT in this release and was withdrawn rather than deferred**: a fix passed eight council dimensions and was then refuted by a five-day-old spike nobody had routed to it — Flow's `/about` hop is decided client-side with zero requests to Flow, and *"the backend grants access while the frontend declines to open it"* (#756), so the probe would have reported authenticated users as logged out, reintroducing #791 as its own cure; the e2e caught it by going red on a live account whose state moved mid-session. **Two rows are blocked by the platform rather than by effort:** the Docker sign-in needs a WSL-integrated Docker host, and the MCP Registry publish cannot be observed until this very tag reaches `main`, since both triggers resolve the workflow from the default branch.) · previous: [LIVE_VERIFICATION_v0.75.0](LIVE_VERIFICATION_v0.75.0.md) (**a distribution-and-hardening release, verified 2026-09-15 on `ffroliva` at ZERO cost — no Veo credits, no image generation, no Flow submit.** The `gflow serve` token fix is an **A/B against the released 0.74.0 wheel from PyPI**, not a diff read: the token was *set* in both arms, so the probe measures whether it is checked — before, `no token 200/200` with 13 tools listed and `wrong 200/200`; after, `401/401` with zero tools and `WWW-Authenticate: Bearer realm="gflow"`, while the correct token still gets 200. Loopback DNS-rebinding was re-confirmed at **421**/**403** because the fix forwards the bind host and getting that wrong would silently disable protection that works today — and an earlier draft of the accompanying advisory claimed that protection was *disabled*, a false claim caught by measuring rather than re-reading. #813's guard is verified against a **real wheel in a real venv installed without extras**, the arm first proving it is in that state (`av present: False  PIL present: False`); the decisive row is that a **nonexistent** manifest still exits **20** on the dependency error rather than file-not-found, which is what shows the guard precedes the manifest read and not merely the generation. The Rich escape fix is verified on the **console render separated from the structlog JSON** — matching the JSON line would have been a false pass, since `--json` was never affected — with a **neutered control** through the same Rich version reproducing the bug exactly: `UNESCAPED -> pip install 'gflow-cli'` against `ESCAPED -> pip install 'gflow-cli[chain]'`. Also: `gflow-cli.exe` exists in the built wheel so `uvx gflow-cli` resolves at all, and the wheel carries 138 entries with **zero** ZIP duplicates. **Three items recorded as NOT verified rather than omitted:** a successful multi-link `chain` run spends Veo credits and stays outstanding e2e debt on #813 — what changed is the refusal path, which *is* verified; `gflow credits` against a working balance is impossible here because the account still returns 401, pre-existing **#795** and unrelated to this release; and the SSE arm of the token fix was exercised on loopback only, a deliberate scope limit, since binding a LAN address to test it would put an authenticated daemon on a live network for the run. **Two rows are blocked by the platform rather than by effort:** the PyPI metadata and the MCP Registry listing are both unobservable until this release is uploaded — PyPI renders the last artefact's metadata, and `mcp-publisher` reads its ownership token out of the *published* README.) · previous: [LIVE_VERIFICATION_v0.74.0](LIVE_VERIFICATION_v0.74.0.md) (**the migrated driver's run-unique uploads and its Frames-picker confirm, plus the `credits` remediation, verified 2026-09-14 on `ffroliva` at ZERO cost — every arm is a read or an upload-and-stop.** The confirm branch is the interesting one: this cohort's picker closes on its own, so the branch would never execute against live Flow by accident, and the e2e drives `FRAME_COMMIT_GRACE_S` to **0.001** so the grace expires before Flow can auto-close and **every** run enters it — the difference between a named blocker and an unrun claim. A *genuinely stuck* picker still cannot be produced here and stays `Refs #792`, verifier the reporter. #795 was verified on the cohort it is for, because this account is now in it — `auth status` green, `credits user` exit **3** with `detail` `credits endpoint returned 401` and `route: credits`, where v0.73.2 printed the class-default *"SAPISID cookie missing… re-run `gflow auth login`"*, advice that is wrong in every word there and that can roll a profile's strategy marker back and start the #791 spiral; the `credits list --json` envelope carries the per-profile `remediation_hint` that used to be dropped. **Two items recorded as NOT verified rather than omitted:** #799's exit-25 raise site needs an account in the agent-only cohort — the maintainer's accounts are served `flow.google.com` but still have the recoverable agent chip, so the state is structurally unreachable on them, a named blocker and #799 stays open; and neither `r2v` nor `i2i` had its own attach path driven to completion — `r2v` was not attempted (its e2e spends Veo credits), while `i2i` **was** attempted twice and failed both times at exit 23, before reaching the code under test, because the `--project` UUID came from `gflow project list` (the aisandbox catalog) and those ids 404 on `flow.google.com` — a harness error, recorded rather than dropped, which left a side finding: gflow reports Flow's own `/404?reason=project` as `UiSelectorDriftError`, blaming our selectors for Flow's 404. **#791 is not fixed and was not attempted blind** — all nine local profiles were probed at $0 and none is in the #791 state (SAPISID + a live `flow.google.com` session + no labs NextAuth token); four print its exact symptom string and none of them is it, because that string is also what a stale profile says.) · previous: [LIVE_VERIFICATION_v0.73.2](LIVE_VERIFICATION_v0.73.2.md) (**the blank incident bundle, reproduced and then cured by a committed e2e test, 2026-09-12 on `ffroliva` at ZERO cost — every arm fails before submit.** The decisive layer is an **A/B run through the same instrument**, not an inspection: on `develop` the new test FAILS in 46.3 s and prints the bundle it got — `tag_counts.div = 0`, `title.length = 0`, `host_category = "other"`, a 4,254-byte screenshot — which is field-for-field the signature [#792](https://github.com/ffroliva/gflow-cli/issues/792) reported from a different account, a different OS and a different command; on the fix it PASSES in 37.8 s with `div` **25**, `title.length` **81**, `host_category` **`flow_app`**, `route` `/404` and a 38.4 KB screenshot. Both arms exit **23** on the same `.settings-trigger-button` drift at `flow.google.com/404?reason=project`, so the arms differ only in what was *captured*, never in what failed. **The reporter's own bundle contained the refutation of their diagnosis**: viewport 1280x720 is `client.py`'s client-owned page, not the 1920x1080 a stray transport tab would have — it was always the driven page, photographed one navigation too late, and the portrait correlation had no mechanism behind it. The existing quality benchmark could not have caught this: it captures deliberately on a **healthy** page, so it graded only bundles that were never at risk. **Recorded as NOT verified rather than omitted:** #796's `PROFILE_MARKER_MISSING` is offline-only because its live trigger is a cookie-store **decryption** failure, which does not occur on Windows/DPAPI — it is the macOS Keychain path of [#768](https://github.com/ffroliva/gflow-cli/issues/768), a **named external blocker**, and the issue stays open; a synthetic attempt to force it failed informatively, since a corrupt store raises `sqlite3.DatabaseError`, bypasses the marker gate and still prints "check network connectivity" — the same misdirection from a different trigger, untouched here. #795's remediation string is unit-asserted only, and the *other* half of #795 — `credits` being labs-only on migrated accounts — is **not fixed**.) · previous: [LIVE_VERIFICATION_v0.73.1](LIVE_VERIFICATION_v0.73.1.md) (**Google's `glue` consent bar covering the migrated composer, measured and then cured, 2026-09-11 on `ci-probe` for one image generation and no Veo credits.** The premise was nearly rejected: both prior sightings of that selector in this repo are labs.google, and the 2026-09-10 spike had read 159 DOM samples on this exact host and project and watched the click land 3/3 — it simply never looked for a cookie bar. Seventeen hours later the bar was up: `position: fixed`, `z-index: 1000`, spanning y 547-655 of a 655 px viewport, with `elementFromPoint` over the settings trigger AND over the image submit both returning `span#glue-cookie-notification-bar-1-label` in **0 of 5** rendered samples hit-testable. **The control arm is what makes it a cure rather than a coincidence**: the same probe re-run through the driver's own `_dismiss_cookie_bar` restored both controls to **4/4**, one attempt, no error. A full `gflow image t2i` then ran end to end on the merged code — `status: ok`, a 807,761-byte `ffd8ffe0` JPEG that Pillow reads as 1376x768 RGB, matching the requested 16:9 and the reply's own dimensions, with `migrated.editor_ready` at 2.0 s and **no** `cookie_bar_not_dismissed`. A side finding the control arm produced by accident: the bar's buttons are `__accept` then `__reject` in DOM order, so anything taking "the first button" accepts cookies for the operator — the shipped fix anchors on reject. **Three items recorded as NOT verified rather than omitted:** the cure against a live bar outside the browser, because this release's own control arm consumed the consent on the only profile that had it; **prevalence**, because the second profile never reached an editor (both arms redirected to `/about`) so it is one account blocked and one already-consented, and KNOWN_ISSUES says so rather than generalising from n=1; and the video path live, which shares the identical function on the identical control but spends Veo credits.) · previous: [LIVE_VERIFICATION_v0.73.0](LIVE_VERIFICATION_v0.73.0.md) (**four error paths that stopped lying about what went wrong, verified 2026-09-10 on `ci-probe` and `denon82` at $0 — no run below reaches a Veo submit.** The decisive arm is an **A/B**, not an inspection: the pre-fix run printed an `accounts.google.com` challenge URL together with the OAuth `state` and `client_id`, and the identical command after the fix gave the same exit 38 and the same landing named with **five secret matches down to zero** — reading the post-fix output alone would only have shown that *these* secrets were absent. A second arm **falsified a claim already written into PR #775's body**: it said the change fixed the RED canary, and the live run showed `denon82` had moved to `accounts.google.com` mid-run, a landing the classifier excluded **by design**, with a unit test encoding the wrong reason — the offline suite was green throughout and could not have caught it. #776's happy path is verified by a real `gflow image t2i` (exit 0, 367 135-byte JPEG, `image_settings_applied` and `prompt_typed` both present, so **three of the four converted click sites** clicked successfully through the new helper); its **failure** path is real Chromium driving a page we wrote, which proves the mechanism and explicitly **not** that Flow still produces that state. **Three items recorded as NOT verified rather than omitted:** the `/about` redirect stopped reproducing (**0/5**, reported as *unmeasured* rather than argued into a retry flag), the `/fx/api/auth/*` landing moved before the run reached it, and #764's *success* path — an external contribution whose error boundary this release verifies but whose auto-selection needs an account currently behind a Google password challenge. The fourth converted click site is the **video** submit, unrun because it spends Veo credits, named rather than implied.) · previous: [LIVE_VERIFICATION_v0.72.0](LIVE_VERIFICATION_v0.72.0.md) (**auto-closing `gflow auth login`, verified 2026-09-09 on a migrated account at $0 across six runs.** Four paths: auto-close fires (`probe=in_context` at `elapsed_s` **84.8**, reproduced 3x), a manual close still verifies from disk (exit 0, never exit 12), `--browser auto` selects and completes, and a window closed **mid-2FA** is noticed in **2 s** (exit 8) where it previously ran the full 600 s deadline with the session endpoint touched **zero** times. **Two runs were discarded rather than counted**, and that is the useful part: the first returned exit 0 at `elapsed_s: 0.2` because a throwaway profile had survived the previous session and nobody signed in — a cached session reading as a clean pass — and three attempts at the manual-close path all returned exit 0 while being the *auto-close* path in disguise, separable only by an **absent** `probe=in_context`. The manual path was obtained by widening the poll interval for one run so the competing branch could not fire; the 2FA path by killing the profile's own Chrome, because hand-timing lost every race on a host that completes passkey sign-in in ~15 s. **Recorded as NOT verified rather than omitted:** the migrated-host image slice (#692) was **re-run this cycle** — CLI `t2i` PASSED first-hand on a moved account, while the queued-MCP `i2i` twin failed with a typed non-retryable exit 27 whose **cause is unknown** — some dialog opened after the file was chosen and no `maseQ` left the page, on an account that has ALREADY consented, so the message's "most likely the one-time upload-terms dialog" does not fit; no incident bundle was written, so nothing names the dialog, and it is filed for a spike rather than guessed; the subprocess fallback is unit-tested only; and the OAuth-callback *mechanism* behind the poll fix is **inferred, not proven** — a host-only gate does not exclude NextAuth's callback, since it runs on the app's own origin, yet all four sign-ins passed under it, so #769 carries the spike that would settle it.) · previous: [LIVE_VERIFICATION_v0.71.1](LIVE_VERIFICATION_v0.71.1.md) (**two migrated-host error paths stopped blaming the wrong thing, 2026-09-08, $0 across three profiles.** The agent-mode recovery landed earlier in this same release (#749) and worked, but its first cut collapsed three outcomes into one message — chip clicked / chip found but click blocked / chip clicked and mode still on — so a modal eating the click sent users to toggle a chip that was never the problem, and genuine selector drift *after* the mode was left was filed under an account setting the driver had already changed. Verified by an e2e that drives the account INTO agent mode and carries its own neutered-selector control; run **twice**, because review fixes landed after the first run and it no longer covered the code. **The larger find:** an account's FIRST upload on this host is blocked by Flow's one-time *"Rights to use this image"* dialog, which renders *after* the file chooser hands the file over — so `_dismiss_dialog` never sees it and the driver spent 60 s waiting for a request the page had already declined to make, then advised re-encoding the image. Six runs on `ci-probe` close the chain: guard fires while the dialog is up → owner-authorised accept → uploads (`media_id 8914400f`) → fresh session uploads clean with `dialog: None`. The order was forced, because the dialog is one-off and accepting it destroys the failing state — so the guard was written and verified BEFORE anything was clicked, and its firing branch is now unrepeatable here. `maseQ` is confirmed **not** renamed (six uploads, three profiles), killing #719's top hypothesis. Also: Flow's queue listing is on **`Zzl0ze`**, not the `jwpduf`/`as29s` progress polls, and two `abra_r2v_8s` records sat at status 6 for ~20 h unseen by any CLI command (#741). Three documents are corrected rather than patched — a KNOWN_ISSUES entry that told #719's reader this dialog did not affect gflow-cli, and a v0.71.0 host label calling `ci-probe` *labs* when it is migrated, which helped a credit-based theory survive four runs. Recorded as NOT verified: #719's second failure shape — an upload request that leaves the page and is never answered, ~1 run in 4 — which is unfixed and keeps #719 open.) · previous: [LIVE_VERIFICATION_v0.71.0](LIVE_VERIFICATION_v0.71.0.md) (**`gflow character create --voice` verified end to end for the first time, 2026-09-07 on profile `ffroliva`.** Before this release a repo-wide grep for `--voice` across `tests/e2e/` matched **nothing** — every voice test was a unit test of the hardcoded `VOICES` constant and the one that looked live parsed a fixture, so a voice that silently failed to attach was invisible to the whole suite while the command exited 0. A live create + `character show` read-back settles a contradiction between two of our own documents: `sent='Charon' stored='Charon' identical=True`, so the Capitalized canonical form round-trips and `CHARACTER_RECON.md`'s "preset id is the lowercased name" does not describe today's wire. Both docs also now record that `personalityNotes` is **Agent-scoped** per Flow's own editor copy, not a control on the audio engine. **A claim that was in the code is overturned:** the `_unported_form` entity guard said the submit "never produces a reply" and that the backend refuses — in fact `MZZa6b` replies with a null payload and the generation is **accepted and queued** (Flow types it `abra_r2v_8s` and renders it); the failure is the *observer*, which waits out `SUBMIT_REPLY_BUDGET_S` (60 s, calibrated on 4.0–4.6 s replies against an idle queue) and exits 9 while the video is still rendering. The guard stays until the observer is fixed, because a timeout reported on a healthy generation is worse than an explicit refusal. Also verified: credit shortfall reports exit **37** with `_raise_if_out_of_credits` called from **both** give-up paths — and it is *short for the selected model*, not empty (the measured account held **50** and asked for `veo-quality`, which costs **100**); the "+ New project" CTA anchored structurally on `add` (Tier-1 matches **1** where every previous entry matched **0**, control 47 ligature nodes) after this same release first asserted the control did not exist and a $0 run disproved it in one click; and the incident-bundle DOM dump de-blinded — it queried `i.google-symbols` only, so **every bundle a migrated user sent carried an empty ligature list**, which is why #727 and #731 stayed invisible. One item recorded as NOT verified rather than omitted: whether a bound character's **voice reaches rendered audio** (#738) — attachment is proven, application is not, because no entity-bound generation has yet returned a file to measure. The blocker is no longer credits or account access but retrieval, and the measurement is pre-calibrated: plate-bound takes of one character gave **88 / 103 / 118 Hz** against a **4.3 Hz** engine noise floor, so an applied Algenib should sit near its sample's **163.3 Hz**.) · previous: [LIVE_VERIFICATION_v0.70.0](LIVE_VERIFICATION_v0.70.0.md) (**reference-to-video on the migrated `flow.google.com` host, verified 2026-09-06 on profile `denon82`.** Three r2v runs bound their references every time; the decisive one is semantic rather than structural — two hand-drawn reference images went in and the clip came back carrying the same marker strokes, face and paper texture, which is what separates "Flow bound the reference" from "Flow accepted and ignored it". A four-beat, two-character piece was then produced end to end and joined with the concat filter: 19.000s, 1280x576, `clip_qa` **ok**, A/V **+0.000s**, both actors recognisable across every cut from one face plate plus the canon repeated verbatim. Also verified as exit-code corrections on the real CLI: `-o <existing directory>` now exits **2 in 0.8s** instead of exit 1 after ~2 min **and a billed, orphaned clip**; `character create` on a moved account exits **36 immediately** instead of a bare `RuntimeError` exit 1 after a 20s wait. Four items recorded as NOT verified rather than omitted: #692's original failure could not be reproduced [the reporter's own re-run was on a build without the fix, so it shows the failure is intermittent, not that the fix works]; Flow CHARACTER entities were never exercised because `character create` cannot run on any account available here; `image`/`scene`/`movie`/`extend` all exit 36 on the migrated host and no unmoved account exists here; and the sign-in-interstitial fidelity question from #699 has no bundle evidence either way.) · previous: [LIVE_VERIFICATION_v0.68.0](LIVE_VERIFICATION_v0.68.0.md) (`gflow update` self-upgrading a real `uv tool` and a real `pipx` install from their own running `gflow.exe` to the PyPI release, Windows 11, zero credits — see its own row below) · previous: [LIVE_VERIFICATION_v0.67.0](LIVE_VERIFICATION_v0.67.0.md) (the migrated `flow.google.com` driver — see its own row below) · previous: [LIVE_VERIFICATION_v0.64.0](LIVE_VERIFICATION_v0.64.0.md) (**#626 `--model omni-flash --end-frame` — first+last interpolation on Omni 1.1 Flash, verified 2026-09-02.** The guard this release removes was evidence-gated, so it was retired on evidence: a route-aborted submit capture on **two distinct Google accounts** (`ffroliva`, `denon82`) fired `video:batchAsyncGenerateVideoStartAndEndImage` with `startImage` AND `endImage` non-null at **zero credits**, ruling out a single-account artifact. The decisive layer is semantic, not structural: the paid 4s generation's **last frame IS the supplied end image** and its first frame is the start image — two visibly different subjects — which is the only check that distinguishes "Flow bound the end frame" from "Flow accepted and ignored it", and which layers 1–4 would pass either way. Render: 4.01s / 720×1280 / h264+aac / 2.4 MB, `MEDIA_GENERATION_STATUS_SUCCESSFUL`. The static capability table is **deleted**, not corrected — it went stale silently once and would again; a post-submit route check (`_assert_i2v_route`) now fails a run whose end frame Flow dropped back to the `StartImage` route, catching a rollback on any account without anyone re-reading a support page. One item is recorded as NOT verified rather than omitted: `--duration 10` + end frame is **submit-verified only** (`duration_set seconds=10` on the correct route with both images bound), because the status poll returned HTTP 401 mid-run and a retry 401'd at `project.create` before submitting — [#561](https://github.com/ffroliva/gflow-cli/issues/561), pre-existing, so verification was stopped rather than spending further credits on an environmental fault.) · previous: [LIVE_VERIFICATION_v0.63.0](LIVE_VERIFICATION_v0.63.0.md) (**`gflow video extend` — continuing a clip past Flow's 8s ceiling, verified 2026-09-01 on profile `ffroliva`, 20 Veo credits.** The decisive observation is that segment 2's `source_media_id` **is** segment 1's `media_id` (`0c9364f3…` seeded from `b9458021…`, then `648f9291…` seeded from `0c9364f3…`) — tail-only chaining, which no mock can establish; had it re-seeded from the source, the output would have been two divergent continuations of the same moment instead of one continuous shot. `extend_model_resolved` logged `candidate_count=99` against `SERVICE_TIER_INTERMEDIATE`, proving the key came from the live capability listing rather than a constant, and plan cost matched actual spend (`20 credits total, balance 875` → `Extended — 2/2 segment(s), 20 credits`). Render: 23.02s / 1280×720 / 24fps / h264+aac. **The run found a defect the offline suite is structurally incapable of seeing:** an extend segment carries **7.000000s** of content, not the 8 Flow advertises and bills, so server-side concat pads every internal seam with a frozen frame and digital silence (15s at −75.1 dB against −29 neighbours, reproduced on a second independent render) — filed in KNOWN_ISSUES with the three questions that must be answered before any clamp, and the reason `--extend N` on t2v/i2v was deliberately NOT shipped. Seven items are recorded as NOT verified rather than omitted: portrait `9:16`, `--aspect` against a mismatched source, chains longer than 2 segments, `--resume-from` against a live partial scene, the insufficient-credits refusal, a live Ctrl+C, and `OperationKind.EXTEND` rows.) · previous: [LIVE_VERIFICATION_v0.62.1](LIVE_VERIFICATION_v0.62.1.md) (**#604 `--model omni-flash` selects again after Flow's `Omni Flash` → `Omni 1.1 Flash` rename, verified 2026-08-30 on profile `ffroliva` at ZERO Veo credits.** The production `_select_video_model` was driven against the real picker: the shipped selector resolves verbatim — `model_selected model=omni_flash via="[role='menuitem']:has-text('Omni'):has-text('Flash'):not(:has-text('[Lower Priority]'))"` — and all six cases came back as expected: the four offered tiers SELECT, a tier Flow does not offer REFUSES, and a deliberately ambiguous selector REFUSES rather than resolving `.first`. Selection happens before submit, so the whole matrix is free. The transport's own refusal diagnostic read the live menu back as `Omni 1.1 Flash / Veo 3.1 - Lite / Fast / Quality`, independently corroborating the fixture. Four items are recorded as NOT verified rather than omitted: the rename *direction* [the `Omni Flash` baseline is a different account, locale and date], an actual `Omni ... [Lower Priority]` entry [Flow has never offered one], a full credit-spending generation on omni-flash, and #539's absence question, which stays open.) · previous: [LIVE_VERIFICATION_v0.62.0](LIVE_VERIFICATION_v0.62.0.md) (**#595 `auto` ≡ `classic` for images, verified 2026-08-28 on profile `ci-probe` — an account Flow had moved to the agentic cohort the day before — at ZERO Veo credits.** With no flags and no env vars the run logged `ui_driver.ui_mode.attempt_exit_agent` → `ui_driver.bound mode=classic ui_mode=classic` and generated a real 768x1376 JPEG (exit 0); the day before, the identical command on the same account bound `auto`→`agentic` and failed. The same log independently closed **#183**: `mode_switch_trigger` and `image_mode_tab`, the selectors it reported as unfindable, both matched. #591 was proven against the real 500-row catalog — **0** occurrences of the string `"None"`, 31 proper JSON `null`s, over a catalog holding 119 NULL-bearing rows. #592 shows exactly one `client.account_locale_cached` per command. Three items are recorded as NOT verified rather than omitted: the exit-28 abort on a *pinned* agentic account, the #597 batch inter-prompt guard, and any fresh announcement modal — the last two because all three accounts have already acked the current changelog, so no modal can be raised until Google ships the next one.) · previous: [LIVE_VERIFICATION_v0.61.0](LIVE_VERIFICATION_v0.61.0.md) (**#539 video-model refusal + #586 image-model refusal and server-side attribution, verified 2026-08-26/27 on profile `denon82`, ONE Veo generation total** — the video fix was proven by a **zero-credit A/B against the stashed pre-fix source running on live Flow**: the old code returned SELECTED for `veo-lite-lp`, a model Flow does not offer to this account, and for a deliberately ambiguous selector matching 3 entries; the new code refuses both with exit 18 naming what Flow offered. Model selection happens before submit, so refusals cost nothing and the A/B was free. The happy path still generates: `--model omni-flash` → exit 0, a real 2.2 MB `ftypisom` mp4, catalog recording `omni_flash`. Two items are recorded as NOT verified rather than omitted: three of #584's four navigation-settle sites were never exercised [no bearer refresh occurred in these runs], and #582's canary self-re-exec has its first real exercise tonight. Also falsified: #539's recorded note that the video picker uses a different trigger — the two constants are byte-identical strings, and the repeated empty menu captures were the capture's own fault.) · previous: [LIVE_VERIFICATION_v0.60.0](LIVE_VERIFICATION_v0.60.0.md)
+**What was actually live-verified for the latest release?** → latest: [LIVE_VERIFICATION_v0.82.1](LIVE_VERIFICATION_v0.82.1.md) (**2026-10-02.** A reCAPTCHA mint failure typed with a live-measured retry flag (#915, e2e 2 passed, $0); a busy catalog no longer fails a successful generation (#900, real SQLite, no Flow surface); a video's workflow id recorded (#898, 1 veo-lite clip)) · previous: [LIVE_VERIFICATION_v0.82.0](LIVE_VERIFICATION_v0.82.0.md) (**2026-10-01.** `"ref": "batch:N"` and local-file refs on `gflow run` verified by a BDD e2e on flow.google.com: in place, no re-upload, lineage recorded; 1 veo-lite clip for the #926 e2e) · previous: [LIVE_VERIFICATION_v0.81.0](LIVE_VERIFICATION_v0.81.0.md) (**2026-09-30, 3 veo-lite clips.** The off-screen generation browser (#923) measured unchanged for image and video, window placement read back from Chrome, focus-steal measured and documented; #919 verified at merge) · previous: [LIVE_VERIFICATION_v0.80.0](LIVE_VERIFICATION_v0.80.0.md) (**2026-09-29, 0 Veo credits.** `--resolution` 360p/720p proven by driving the real settings pane on `omni_flash` and refusing the credit-spending submit, on the CLI transport **and** the MCP twin; `nano2-lite` e2e 2 passed; #891 measured before/after — `gflow run` `[nano2, imagen4]` went from exit 1 `RecaptchaError` with a lost results table to exit 36 naming the model, plus 3 e2e (fresh, warm, MCP) with a spy asserting zero client mints; #896 `e2e_data` 2 passed, its failure path offline-only. Named blockers: rendered resolution (needs a billed clip), the `/about` re-check arm, labs-served accounts.) · previous: [LIVE_VERIFICATION_v0.79.1](LIVE_VERIFICATION_v0.79.1.md) (**the signed-media download survives a dropped connection, verified 2026-09-22 at $0.** All three download sites route through one new shared helper, and that helper was driven through a real Flow clip route and a real signed CDN transfer by `pytest -m e2e_data` — 2 passed in 20.74 s, `ftyp` magic, byte length equal to the size Flow reports, and **no `url=` on any log event**. The retry itself could not wait for an organic reset — 20/20 clean here against the reporter's 2-in-6 — so it was **measured against a real TCP RST with an A/B control**: with `max_retries` the transfer recovers on a second connection, with it disabled the identical fault kills the call, and three attempts at a 5 s timeout finish in 1.57 s rather than ~15 s, proving the backoff shares one budget. **Two sites are NOT live-verified and name their blockers:** the composer call site needs a billed generation and `gflow credits` answers 401 on this account (#795), so no credits were spent; and the labs download cannot be reached at all from here (HTTP 308, 6/6 profiles). Offline: 21 new tests, full sweep 2143 passed.) · earlier: [LIVE_VERIFICATION_v0.79.0](LIVE_VERIFICATION_v0.79.0.md) (**the migrated host stops echoing a model it never observed, verified 2026-09-18 at $0.** `model_name_type` is `null` on `flow.google.com` in a real `image t2i --json`, with every sibling field populated — which is what makes the null a result rather than an empty envelope; the first read of it was flat rather than into `images[]` and returned `None` for everything, and that instrument error was discarded rather than reported. Both surfaces ran: `test_e2e_t2i_runs_on_a_moved_account` and `test_e2e_mcp_i2i_runs_on_the_migrated_host`, 2 passed in 86 s. **`gflow data download` was tested on an image and failed** — 45 s, then a message blaming the project, the trash and your prompt; recovery is video-only and the release notes now say so (#877). **#803 is NOT live-verified and names its blocker:** reaching those raise sites needs a non-JSON or token-less aisandbox reply, and the labs route refuses first with a 404 — itself a misleading error, filed as #875. Offline: 4 new unit tests, 1775 passed.) · previous: [LIVE_VERIFICATION_v0.78.0](LIVE_VERIFICATION_v0.78.0.md) (**generating with no project on `flow.google.com` — and video from a start frame — verified 2026-09-17.** The release exists because labs.google's `project.createProject` is retired — 404 *"Flow RPCs have been deprecated and disabled"* with a labs token, 401 without, 0 successes in 12 across four profiles — and gflow reported the 401 as an expired login. Creation and rename are driven on flow.google.com and read back from Flow's own `jHPbke` / `o8DA4` replies; the CLI and the MCP twin are each exercised separately, and both video paths reach a real submit that is aborted before Flow can bill it. Two failures are recorded and re-tested rather than explained away — one MCP image reply that did not reproduce under a four-run A/B, and a Chrome launch starved by a concurrent coverage run that passed in 36 s alone. The 3:4 image aspect is re-measured on the live radiogroup, five radios where 2026-09-08 saw four. Three items are NOT verified and each names its blocker: no labs-served account, no entitled `/about` account, no non-`en` locale. **#860 is the one thing here that cost credits, deliberately:** the only proof that a start frame binds is a real generation, so ten i2v clips were generated and downloaded on a live migrated account — `migrated.frame_uploaded` → `frame_bound` → `submit_observed` → `download`, the middle event being the one that could not fire before this release. **#859 is recorded as UNOBSERVED, not verified** — its dismissal ran on every one of those runs and no promo modal ever appeared, so the $0 route-intercepted e2e is what covers it. `gflow docs` (#861) is verified from the built wheel, which is its whole contract: 127 topics, `docs/FILE.md:LINE` search hits, no network and no checkout.) · previous: [LIVE_VERIFICATION_v0.77.1](LIVE_VERIFICATION_v0.77.1.md) (**a four-fix patch release, verified 2026-09-17 at ZERO cost.** The headline is #846: the published package could not run a single interactive command on a clean Windows install, because `ConsoleRenderer(colors=True)` needs `colorama` and nothing declared it. Verified as a controlled A/B on a fresh Python 3.13 venv — `SystemError` before, exit 33 after — plus an arm with stderr piped that shows why the AUTO/JSON split hid this from every CI job, and the MCP twin run separately. #848's import probe is exercised against a **deliberately broken venv** and an intact one, including the contract that a probe which cannot run reports nothing rather than reporting breakage. #852 is re-measured on the release head: 40 000 base64-shaped chars fall from 12.07 s to 0.020 ms with identical output on Workspace and custom domains. **#849 is a NAMED BLOCKER, not an omission** — its end-to-end behaviour needs a fresh interactive Google sign-in, which no agent can perform; its live A/B is recorded on PR #851 and its e2e file re-ran 4/4 here. **Seven live e2e failures are recorded rather than hidden**, and controlled: the release changes exactly four source files and none of them is a failing test's module.) · previous: [LIVE_VERIFICATION_v0.77.0](LIVE_VERIFICATION_v0.77.0.md) (**the migrated-host login lockout (#791), verified as a controlled A/B on a real Google **Workspace** account at ZERO cost — same profile, same cookies untouched between arms, so the only variable is the code: exit 8 with the Chrome marker rolled back and `channel_for_profile` collapsed to `None`, then exit 0 with the marker intact and the account recorded. The refusal path is measured too — of 10 local profiles, 5 reproduce #791 and exactly one carries both required cookies: 4 refusals, 1 upgrade, 0 false positives. v0.76.0 recorded this fix as **withdrawn**; the approach that shipped is a different one, server-attested on the landing URL rather than on client-rendered DOM. An earlier revision gated auth on an `@gmail.com`-only pattern that silently declined every custom domain — the account used here is one, which is why this is its verification. **#796 is reproduced and prevented**, not merely reported: it was filed as *'not yet reproduced end-to-end'*. **Two rows are named blockers, not omissions:** the MCP Registry publish (#841) cannot be rehearsed before a real tag push, so this release is its first exercise; and the migrated-host generation e2e cannot run here at all — every account bounces to `/about`, A/B-controlled against `develop` so it is not mistaken for a regression.) · previous: [LIVE_VERIFICATION_v0.76.0](LIVE_VERIFICATION_v0.76.0.md) (**a diagnostics-and-distribution release, verified 2026-09-16 at ZERO cost — no Veo credits, no image generation, no Flow submit.** The decisive arm is #833's **negative** control, run first-hand on two real accounts rather than relayed: `flow-pinhole-unavailable-screen` is **0** on both `ffroliva` and `denon82` while `aisandbox-root` is **1**, so the guard does not fire on accounts that HAVE Flow access and the zero is a real absence rather than a page that never booted — a guard with only a positive arm is indistinguishable from one that is permanently on. Both accounts settled at **`/about`**, which is the state that broke this release's *withdrawn* #791 oracle, and the exit-39 signal reads a component that only exists on the unavailable screen, so `/about` cannot mimic it. mcpservers.org is verified with a **404 negative control** because the listing 403s to `curl` and a naive check would have called the link dead. **#791 is NOT in this release and was withdrawn rather than deferred**: a fix passed eight council dimensions and was then refuted by a five-day-old spike nobody had routed to it — Flow's `/about` hop is decided client-side with zero requests to Flow, and *"the backend grants access while the frontend declines to open it"* (#756), so the probe would have reported authenticated users as logged out, reintroducing #791 as its own cure; the e2e caught it by going red on a live account whose state moved mid-session. **Two rows are blocked by the platform rather than by effort:** the Docker sign-in needs a WSL-integrated Docker host, and the MCP Registry publish cannot be observed until this very tag reaches `main`, since both triggers resolve the workflow from the default branch.) · previous: [LIVE_VERIFICATION_v0.75.0](LIVE_VERIFICATION_v0.75.0.md) (**a distribution-and-hardening release, verified 2026-09-15 on `ffroliva` at ZERO cost — no Veo credits, no image generation, no Flow submit.** The `gflow serve` token fix is an **A/B against the released 0.74.0 wheel from PyPI**, not a diff read: the token was *set* in both arms, so the probe measures whether it is checked — before, `no token 200/200` with 13 tools listed and `wrong 200/200`; after, `401/401` with zero tools and `WWW-Authenticate: Bearer realm="gflow"`, while the correct token still gets 200. Loopback DNS-rebinding was re-confirmed at **421**/**403** because the fix forwards the bind host and getting that wrong would silently disable protection that works today — and an earlier draft of the accompanying advisory claimed that protection was *disabled*, a false claim caught by measuring rather than re-reading. #813's guard is verified against a **real wheel in a real venv installed without extras**, the arm first proving it is in that state (`av present: False  PIL present: False`); the decisive row is that a **nonexistent** manifest still exits **20** on the dependency error rather than file-not-found, which is what shows the guard precedes the manifest read and not merely the generation. The Rich escape fix is verified on the **console render separated from the structlog JSON** — matching the JSON line would have been a false pass, since `--json` was never affected — with a **neutered control** through the same Rich version reproducing the bug exactly: `UNESCAPED -> pip install 'gflow-cli'` against `ESCAPED -> pip install 'gflow-cli[chain]'`. Also: `gflow-cli.exe` exists in the built wheel so `uvx gflow-cli` resolves at all, and the wheel carries 138 entries with **zero** ZIP duplicates. **Three items recorded as NOT verified rather than omitted:** a successful multi-link `chain` run spends Veo credits and stays outstanding e2e debt on #813 — what changed is the refusal path, which *is* verified; `gflow credits` against a working balance is impossible here because the account still returns 401, pre-existing **#795** and unrelated to this release; and the SSE arm of the token fix was exercised on loopback only, a deliberate scope limit, since binding a LAN address to test it would put an authenticated daemon on a live network for the run. **Two rows are blocked by the platform rather than by effort:** the PyPI metadata and the MCP Registry listing are both unobservable until this release is uploaded — PyPI renders the last artefact's metadata, and `mcp-publisher` reads its ownership token out of the *published* README.) · previous: [LIVE_VERIFICATION_v0.74.0](LIVE_VERIFICATION_v0.74.0.md) (**the migrated driver's run-unique uploads and its Frames-picker confirm, plus the `credits` remediation, verified 2026-09-14 on `ffroliva` at ZERO cost — every arm is a read or an upload-and-stop.** The confirm branch is the interesting one: this cohort's picker closes on its own, so the branch would never execute against live Flow by accident, and the e2e drives `FRAME_COMMIT_GRACE_S` to **0.001** so the grace expires before Flow can auto-close and **every** run enters it — the difference between a named blocker and an unrun claim. A *genuinely stuck* picker still cannot be produced here and stays `Refs #792`, verifier the reporter. #795 was verified on the cohort it is for, because this account is now in it — `auth status` green, `credits user` exit **3** with `detail` `credits endpoint returned 401` and `route: credits`, where v0.73.2 printed the class-default *"SAPISID cookie missing… re-run `gflow auth login`"*, advice that is wrong in every word there and that can roll a profile's strategy marker back and start the #791 spiral; the `credits list --json` envelope carries the per-profile `remediation_hint` that used to be dropped. **Two items recorded as NOT verified rather than omitted:** #799's exit-25 raise site needs an account in the agent-only cohort — the maintainer's accounts are served `flow.google.com` but still have the recoverable agent chip, so the state is structurally unreachable on them, a named blocker and #799 stays open; and neither `r2v` nor `i2i` had its own attach path driven to completion — `r2v` was not attempted (its e2e spends Veo credits), while `i2i` **was** attempted twice and failed both times at exit 23, before reaching the code under test, because the `--project` UUID came from `gflow project list` (the aisandbox catalog) and those ids 404 on `flow.google.com` — a harness error, recorded rather than dropped, which left a side finding: gflow reports Flow's own `/404?reason=project` as `UiSelectorDriftError`, blaming our selectors for Flow's 404. **#791 is not fixed and was not attempted blind** — all nine local profiles were probed at $0 and none is in the #791 state (SAPISID + a live `flow.google.com` session + no labs NextAuth token); four print its exact symptom string and none of them is it, because that string is also what a stale profile says.) · previous: [LIVE_VERIFICATION_v0.73.2](LIVE_VERIFICATION_v0.73.2.md) (**the blank incident bundle, reproduced and then cured by a committed e2e test, 2026-09-12 on `ffroliva` at ZERO cost — every arm fails before submit.** The decisive layer is an **A/B run through the same instrument**, not an inspection: on `develop` the new test FAILS in 46.3 s and prints the bundle it got — `tag_counts.div = 0`, `title.length = 0`, `host_category = "other"`, a 4,254-byte screenshot — which is field-for-field the signature [#792](https://github.com/ffroliva/gflow-cli/issues/792) reported from a different account, a different OS and a different command; on the fix it PASSES in 37.8 s with `div` **25**, `title.length` **81**, `host_category` **`flow_app`**, `route` `/404` and a 38.4 KB screenshot. Both arms exit **23** on the same `.settings-trigger-button` drift at `flow.google.com/404?reason=project`, so the arms differ only in what was *captured*, never in what failed. **The reporter's own bundle contained the refutation of their diagnosis**: viewport 1280x720 is `client.py`'s client-owned page, not the 1920x1080 a stray transport tab would have — it was always the driven page, photographed one navigation too late, and the portrait correlation had no mechanism behind it. The existing quality benchmark could not have caught this: it captures deliberately on a **healthy** page, so it graded only bundles that were never at risk. **Recorded as NOT verified rather than omitted:** #796's `PROFILE_MARKER_MISSING` is offline-only because its live trigger is a cookie-store **decryption** failure, which does not occur on Windows/DPAPI — it is the macOS Keychain path of [#768](https://github.com/ffroliva/gflow-cli/issues/768), a **named external blocker**, and the issue stays open; a synthetic attempt to force it failed informatively, since a corrupt store raises `sqlite3.DatabaseError`, bypasses the marker gate and still prints "check network connectivity" — the same misdirection from a different trigger, untouched here. #795's remediation string is unit-asserted only, and the *other* half of #795 — `credits` being labs-only on migrated accounts — is **not fixed**.) · previous: [LIVE_VERIFICATION_v0.73.1](LIVE_VERIFICATION_v0.73.1.md) (**Google's `glue` consent bar covering the migrated composer, measured and then cured, 2026-09-11 on `ci-probe` for one image generation and no Veo credits.** The premise was nearly rejected: both prior sightings of that selector in this repo are labs.google, and the 2026-09-10 spike had read 159 DOM samples on this exact host and project and watched the click land 3/3 — it simply never looked for a cookie bar. Seventeen hours later the bar was up: `position: fixed`, `z-index: 1000`, spanning y 547-655 of a 655 px viewport, with `elementFromPoint` over the settings trigger AND over the image submit both returning `span#glue-cookie-notification-bar-1-label` in **0 of 5** rendered samples hit-testable. **The control arm is what makes it a cure rather than a coincidence**: the same probe re-run through the driver's own `_dismiss_cookie_bar` restored both controls to **4/4**, one attempt, no error. A full `gflow image t2i` then ran end to end on the merged code — `status: ok`, a 807,761-byte `ffd8ffe0` JPEG that Pillow reads as 1376x768 RGB, matching the requested 16:9 and the reply's own dimensions, with `migrated.editor_ready` at 2.0 s and **no** `cookie_bar_not_dismissed`. A side finding the control arm produced by accident: the bar's buttons are `__accept` then `__reject` in DOM order, so anything taking "the first button" accepts cookies for the operator — the shipped fix anchors on reject. **Three items recorded as NOT verified rather than omitted:** the cure against a live bar outside the browser, because this release's own control arm consumed the consent on the only profile that had it; **prevalence**, because the second profile never reached an editor (both arms redirected to `/about`) so it is one account blocked and one already-consented, and KNOWN_ISSUES says so rather than generalising from n=1; and the video path live, which shares the identical function on the identical control but spends Veo credits.) · previous: [LIVE_VERIFICATION_v0.73.0](LIVE_VERIFICATION_v0.73.0.md) (**four error paths that stopped lying about what went wrong, verified 2026-09-10 on `ci-probe` and `denon82` at $0 — no run below reaches a Veo submit.** The decisive arm is an **A/B**, not an inspection: the pre-fix run printed an `accounts.google.com` challenge URL together with the OAuth `state` and `client_id`, and the identical command after the fix gave the same exit 38 and the same landing named with **five secret matches down to zero** — reading the post-fix output alone would only have shown that *these* secrets were absent. A second arm **falsified a claim already written into PR #775's body**: it said the change fixed the RED canary, and the live run showed `denon82` had moved to `accounts.google.com` mid-run, a landing the classifier excluded **by design**, with a unit test encoding the wrong reason — the offline suite was green throughout and could not have caught it. #776's happy path is verified by a real `gflow image t2i` (exit 0, 367 135-byte JPEG, `image_settings_applied` and `prompt_typed` both present, so **three of the four converted click sites** clicked successfully through the new helper); its **failure** path is real Chromium driving a page we wrote, which proves the mechanism and explicitly **not** that Flow still produces that state. **Three items recorded as NOT verified rather than omitted:** the `/about` redirect stopped reproducing (**0/5**, reported as *unmeasured* rather than argued into a retry flag), the `/fx/api/auth/*` landing moved before the run reached it, and #764's *success* path — an external contribution whose error boundary this release verifies but whose auto-selection needs an account currently behind a Google password challenge. The fourth converted click site is the **video** submit, unrun because it spends Veo credits, named rather than implied.) · previous: [LIVE_VERIFICATION_v0.72.0](LIVE_VERIFICATION_v0.72.0.md) (**auto-closing `gflow auth login`, verified 2026-09-09 on a migrated account at $0 across six runs.** Four paths: auto-close fires (`probe=in_context` at `elapsed_s` **84.8**, reproduced 3x), a manual close still verifies from disk (exit 0, never exit 12), `--browser auto` selects and completes, and a window closed **mid-2FA** is noticed in **2 s** (exit 8) where it previously ran the full 600 s deadline with the session endpoint touched **zero** times. **Two runs were discarded rather than counted**, and that is the useful part: the first returned exit 0 at `elapsed_s: 0.2` because a throwaway profile had survived the previous session and nobody signed in — a cached session reading as a clean pass — and three attempts at the manual-close path all returned exit 0 while being the *auto-close* path in disguise, separable only by an **absent** `probe=in_context`. The manual path was obtained by widening the poll interval for one run so the competing branch could not fire; the 2FA path by killing the profile's own Chrome, because hand-timing lost every race on a host that completes passkey sign-in in ~15 s. **Recorded as NOT verified rather than omitted:** the migrated-host image slice (#692) was **re-run this cycle** — CLI `t2i` PASSED first-hand on a moved account, while the queued-MCP `i2i` twin failed with a typed non-retryable exit 27 whose **cause is unknown** — some dialog opened after the file was chosen and no `maseQ` left the page, on an account that has ALREADY consented, so the message's "most likely the one-time upload-terms dialog" does not fit; no incident bundle was written, so nothing names the dialog, and it is filed for a spike rather than guessed; the subprocess fallback is unit-tested only; and the OAuth-callback *mechanism* behind the poll fix is **inferred, not proven** — a host-only gate does not exclude NextAuth's callback, since it runs on the app's own origin, yet all four sign-ins passed under it, so #769 carries the spike that would settle it.) · previous: [LIVE_VERIFICATION_v0.71.1](LIVE_VERIFICATION_v0.71.1.md) (**two migrated-host error paths stopped blaming the wrong thing, 2026-09-08, $0 across three profiles.** The agent-mode recovery landed earlier in this same release (#749) and worked, but its first cut collapsed three outcomes into one message — chip clicked / chip found but click blocked / chip clicked and mode still on — so a modal eating the click sent users to toggle a chip that was never the problem, and genuine selector drift *after* the mode was left was filed under an account setting the driver had already changed. Verified by an e2e that drives the account INTO agent mode and carries its own neutered-selector control; run **twice**, because review fixes landed after the first run and it no longer covered the code. **The larger find:** an account's FIRST upload on this host is blocked by Flow's one-time *"Rights to use this image"* dialog, which renders *after* the file chooser hands the file over — so `_dismiss_dialog` never sees it and the driver spent 60 s waiting for a request the page had already declined to make, then advised re-encoding the image. Six runs on `ci-probe` close the chain: guard fires while the dialog is up → owner-authorised accept → uploads (`media_id 8914400f`) → fresh session uploads clean with `dialog: None`. The order was forced, because the dialog is one-off and accepting it destroys the failing state — so the guard was written and verified BEFORE anything was clicked, and its firing branch is now unrepeatable here. `maseQ` is confirmed **not** renamed (six uploads, three profiles), killing #719's top hypothesis. Also: Flow's queue listing is on **`Zzl0ze`**, not the `jwpduf`/`as29s` progress polls, and two `abra_r2v_8s` records sat at status 6 for ~20 h unseen by any CLI command (#741). Three documents are corrected rather than patched — a KNOWN_ISSUES entry that told #719's reader this dialog did not affect gflow-cli, and a v0.71.0 host label calling `ci-probe` *labs* when it is migrated, which helped a credit-based theory survive four runs. Recorded as NOT verified: #719's second failure shape — an upload request that leaves the page and is never answered, ~1 run in 4 — which is unfixed and keeps #719 open.) · previous: [LIVE_VERIFICATION_v0.71.0](LIVE_VERIFICATION_v0.71.0.md) (**`gflow character create --voice` verified end to end for the first time, 2026-09-07 on profile `ffroliva`.** Before this release a repo-wide grep for `--voice` across `tests/e2e/` matched **nothing** — every voice test was a unit test of the hardcoded `VOICES` constant and the one that looked live parsed a fixture, so a voice that silently failed to attach was invisible to the whole suite while the command exited 0. A live create + `character show` read-back settles a contradiction between two of our own documents: `sent='Charon' stored='Charon' identical=True`, so the Capitalized canonical form round-trips and `CHARACTER_RECON.md`'s "preset id is the lowercased name" does not describe today's wire. Both docs also now record that `personalityNotes` is **Agent-scoped** per Flow's own editor copy, not a control on the audio engine. **A claim that was in the code is overturned:** the `_unported_form` entity guard said the submit "never produces a reply" and that the backend refuses — in fact `MZZa6b` replies with a null payload and the generation is **accepted and queued** (Flow types it `abra_r2v_8s` and renders it); the failure is the *observer*, which waits out `SUBMIT_REPLY_BUDGET_S` (60 s, calibrated on 4.0–4.6 s replies against an idle queue) and exits 9 while the video is still rendering. The guard stays until the observer is fixed, because a timeout reported on a healthy generation is worse than an explicit refusal. Also verified: credit shortfall reports exit **37** with `_raise_if_out_of_credits` called from **both** give-up paths — and it is *short for the selected model*, not empty (the measured account held **50** and asked for `veo-quality`, which costs **100**); the "+ New project" CTA anchored structurally on `add` (Tier-1 matches **1** where every previous entry matched **0**, control 47 ligature nodes) after this same release first asserted the control did not exist and a $0 run disproved it in one click; and the incident-bundle DOM dump de-blinded — it queried `i.google-symbols` only, so **every bundle a migrated user sent carried an empty ligature list**, which is why #727 and #731 stayed invisible. One item recorded as NOT verified rather than omitted: whether a bound character's **voice reaches rendered audio** (#738) — attachment is proven, application is not, because no entity-bound generation has yet returned a file to measure. The blocker is no longer credits or account access but retrieval, and the measurement is pre-calibrated: plate-bound takes of one character gave **88 / 103 / 118 Hz** against a **4.3 Hz** engine noise floor, so an applied Algenib should sit near its sample's **163.3 Hz**.) · previous: [LIVE_VERIFICATION_v0.70.0](LIVE_VERIFICATION_v0.70.0.md) (**reference-to-video on the migrated `flow.google.com` host, verified 2026-09-06 on profile `denon82`.** Three r2v runs bound their references every time; the decisive one is semantic rather than structural — two hand-drawn reference images went in and the clip came back carrying the same marker strokes, face and paper texture, which is what separates "Flow bound the reference" from "Flow accepted and ignored it". A four-beat, two-character piece was then produced end to end and joined with the concat filter: 19.000s, 1280x576, `clip_qa` **ok**, A/V **+0.000s**, both actors recognisable across every cut from one face plate plus the canon repeated verbatim. Also verified as exit-code corrections on the real CLI: `-o <existing directory>` now exits **2 in 0.8s** instead of exit 1 after ~2 min **and a billed, orphaned clip**; `character create` on a moved account exits **36 immediately** instead of a bare `RuntimeError` exit 1 after a 20s wait. Four items recorded as NOT verified rather than omitted: #692's original failure could not be reproduced [the reporter's own re-run was on a build without the fix, so it shows the failure is intermittent, not that the fix works]; Flow CHARACTER entities were never exercised because `character create` cannot run on any account available here; `image`/`scene`/`movie`/`extend` all exit 36 on the migrated host and no unmoved account exists here; and the sign-in-interstitial fidelity question from #699 has no bundle evidence either way.) · previous: [LIVE_VERIFICATION_v0.68.0](LIVE_VERIFICATION_v0.68.0.md) (`gflow update` self-upgrading a real `uv tool` and a real `pipx` install from their own running `gflow.exe` to the PyPI release, Windows 11, zero credits — see its own row below) · previous: [LIVE_VERIFICATION_v0.67.0](LIVE_VERIFICATION_v0.67.0.md) (the migrated `flow.google.com` driver — see its own row below) · previous: [LIVE_VERIFICATION_v0.64.0](LIVE_VERIFICATION_v0.64.0.md) (**#626 `--model omni-flash --end-frame` — first+last interpolation on Omni 1.1 Flash, verified 2026-09-02.** The guard this release removes was evidence-gated, so it was retired on evidence: a route-aborted submit capture on **two distinct Google accounts** (`ffroliva`, `denon82`) fired `video:batchAsyncGenerateVideoStartAndEndImage` with `startImage` AND `endImage` non-null at **zero credits**, ruling out a single-account artifact. The decisive layer is semantic, not structural: the paid 4s generation's **last frame IS the supplied end image** and its first frame is the start image — two visibly different subjects — which is the only check that distinguishes "Flow bound the end frame" from "Flow accepted and ignored it", and which layers 1–4 would pass either way. Render: 4.01s / 720×1280 / h264+aac / 2.4 MB, `MEDIA_GENERATION_STATUS_SUCCESSFUL`. The static capability table is **deleted**, not corrected — it went stale silently once and would again; a post-submit route check (`_assert_i2v_route`) now fails a run whose end frame Flow dropped back to the `StartImage` route, catching a rollback on any account without anyone re-reading a support page. One item is recorded as NOT verified rather than omitted: `--duration 10` + end frame is **submit-verified only** (`duration_set seconds=10` on the correct route with both images bound), because the status poll returned HTTP 401 mid-run and a retry 401'd at `project.create` before submitting — [#561](https://github.com/ffroliva/gflow-cli/issues/561), pre-existing, so verification was stopped rather than spending further credits on an environmental fault.) · previous: [LIVE_VERIFICATION_v0.63.0](LIVE_VERIFICATION_v0.63.0.md) (**`gflow video extend` — continuing a clip past Flow's 8s ceiling, verified 2026-09-01 on profile `ffroliva`, 20 Veo credits.** The decisive observation is that segment 2's `source_media_id` **is** segment 1's `media_id` (`0c9364f3…` seeded from `b9458021…`, then `648f9291…` seeded from `0c9364f3…`) — tail-only chaining, which no mock can establish; had it re-seeded from the source, the output would have been two divergent continuations of the same moment instead of one continuous shot. `extend_model_resolved` logged `candidate_count=99` against `SERVICE_TIER_INTERMEDIATE`, proving the key came from the live capability listing rather than a constant, and plan cost matched actual spend (`20 credits total, balance 875` → `Extended — 2/2 segment(s), 20 credits`). Render: 23.02s / 1280×720 / 24fps / h264+aac. **The run found a defect the offline suite is structurally incapable of seeing:** an extend segment carries **7.000000s** of content, not the 8 Flow advertises and bills, so server-side concat pads every internal seam with a frozen frame and digital silence (15s at −75.1 dB against −29 neighbours, reproduced on a second independent render) — filed in KNOWN_ISSUES with the three questions that must be answered before any clamp, and the reason `--extend N` on t2v/i2v was deliberately NOT shipped. Seven items are recorded as NOT verified rather than omitted: portrait `9:16`, `--aspect` against a mismatched source, chains longer than 2 segments, `--resume-from` against a live partial scene, the insufficient-credits refusal, a live Ctrl+C, and `OperationKind.EXTEND` rows.) · previous: [LIVE_VERIFICATION_v0.62.1](LIVE_VERIFICATION_v0.62.1.md) (**#604 `--model omni-flash` selects again after Flow's `Omni Flash` → `Omni 1.1 Flash` rename, verified 2026-08-30 on profile `ffroliva` at ZERO Veo credits.** The production `_select_video_model` was driven against the real picker: the shipped selector resolves verbatim — `model_selected model=omni_flash via="[role='menuitem']:has-text('Omni'):has-text('Flash'):not(:has-text('[Lower Priority]'))"` — and all six cases came back as expected: the four offered tiers SELECT, a tier Flow does not offer REFUSES, and a deliberately ambiguous selector REFUSES rather than resolving `.first`. Selection happens before submit, so the whole matrix is free. The transport's own refusal diagnostic read the live menu back as `Omni 1.1 Flash / Veo 3.1 - Lite / Fast / Quality`, independently corroborating the fixture. Four items are recorded as NOT verified rather than omitted: the rename *direction* [the `Omni Flash` baseline is a different account, locale and date], an actual `Omni ... [Lower Priority]` entry [Flow has never offered one], a full credit-spending generation on omni-flash, and #539's absence question, which stays open.) · previous: [LIVE_VERIFICATION_v0.62.0](LIVE_VERIFICATION_v0.62.0.md) (**#595 `auto` ≡ `classic` for images, verified 2026-08-28 on profile `ci-probe` — an account Flow had moved to the agentic cohort the day before — at ZERO Veo credits.** With no flags and no env vars the run logged `ui_driver.ui_mode.attempt_exit_agent` → `ui_driver.bound mode=classic ui_mode=classic` and generated a real 768x1376 JPEG (exit 0); the day before, the identical command on the same account bound `auto`→`agentic` and failed. The same log independently closed **#183**: `mode_switch_trigger` and `image_mode_tab`, the selectors it reported as unfindable, both matched. #591 was proven against the real 500-row catalog — **0** occurrences of the string `"None"`, 31 proper JSON `null`s, over a catalog holding 119 NULL-bearing rows. #592 shows exactly one `client.account_locale_cached` per command. Three items are recorded as NOT verified rather than omitted: the exit-28 abort on a *pinned* agentic account, the #597 batch inter-prompt guard, and any fresh announcement modal — the last two because all three accounts have already acked the current changelog, so no modal can be raised until Google ships the next one.) · previous: [LIVE_VERIFICATION_v0.61.0](LIVE_VERIFICATION_v0.61.0.md) (**#539 video-model refusal + #586 image-model refusal and server-side attribution, verified 2026-08-26/27 on profile `denon82`, ONE Veo generation total** — the video fix was proven by a **zero-credit A/B against the stashed pre-fix source running on live Flow**: the old code returned SELECTED for `veo-lite-lp`, a model Flow does not offer to this account, and for a deliberately ambiguous selector matching 3 entries; the new code refuses both with exit 18 naming what Flow offered. Model selection happens before submit, so refusals cost nothing and the A/B was free. The happy path still generates: `--model omni-flash` → exit 0, a real 2.2 MB `ftypisom` mp4, catalog recording `omni_flash`. Two items are recorded as NOT verified rather than omitted: three of #584's four navigation-settle sites were never exercised [no bearer refresh occurred in these runs], and #582's canary self-re-exec has its first real exercise tonight. Also falsified: #539's recorded note that the video picker uses a different trigger — the two constants are byte-identical strings, and the repeated empty menu captures were the capture's own fault.) · previous: [LIVE_VERIFICATION_v0.60.0](LIVE_VERIFICATION_v0.60.0.md)
 **"Was the playwright upper bound raised, and on what evidence?"** → [LIVE_VERIFICATION_playwright_1.61](LIVE_VERIFICATION_playwright_1.61.md) (**`>=1.59.0,<1.60.0` → `>=1.61.0,<1.62.0` verified 2026-08-05, 1 Imagen + 1 Veo credit** — the 2026-08-03 regression that motivated the bound does NOT reproduce on 1.61.0: a live `i2v` drove the full chain with `image_uploaded status=200` → `frame_attached` → `generate_captured status=200` with `startImage` parsed, where **1.62.0 hung silently at exactly that upload step**; live `i2i` local-ref attach passed outright [96.7s]. **1.62.0 stays excluded** — never root-caused. Also found, A/B-proven pre-existing on BOTH 1.59.0 and 1.61.0: Flow dropped the duration-tab UI, so **`--duration` is currently broken for i2v** — `UiSelectorDriftError` fires correctly [refusing rather than silently accepting Flow's default, #288] but the selector needs re-deriving; that is why the i2v proof was driven via the CLI without `--duration`, and Flow's own server-side `PUBLIC_ERROR_VIDEO_GENERATION_TIMED_OUT` means no finished mp4 is claimed).
 **"Where is the reverse-engineered wire protocol for a feature?"** → the `*_RECON.md` design docs: [CHARACTER_RECON](CHARACTER_RECON.md) (Flow character entity protocol), [IMAGE_UPSCALE_RECON](IMAGE_UPSCALE_RECON.md) (`/v1/flow/upsampleImage` wire). Naming convention: one `<FEATURE>_RECON.md` per reverse-engineered surface, kept as the durable spec after the feature ships. Pre-capture research recons: [ASSET_TAGGING_RECON](ASSET_TAGGING_RECON.md) (`@`-mention asset tagging — shipped in v0.40.0, see [LIVE_VERIFICATION_v0.40.0](LIVE_VERIFICATION_v0.40.0.md)).
 **"What was live-verified for v0.66.1?"** → [LIVE_VERIFICATION_v0.66.1](LIVE_VERIFICATION_v0.66.1.md) — **the two migrated-origin fixes, verified 2026-09-03 on profile `ffroliva` at ZERO credits, on BOTH sides of what was then read as a flapping rollout** (superseded 2026-09-04: the handoff is a one-way per-account flag and exit 36 is non-retryable since). On a real `flow.google.com` load `get_ui_driver` raised `FlowHostMigratedError` in **0 ms** (was ~36 s: ~8 s detect window + ~24 s crop cascade + 4 s URL settle), exit 36, `retryable: true` (then; `false` since 2026-09-04), `ui_driver.migrated_host_bail` logged; `await_url_settled` returned `null` in **0 ms** (measured 4018 ms before); and the locale was recovered as `en` from `html lang=en-GB` where the URL gave `null`. The no-regression half is the stronger evidence: minutes later the same command landed on the OLD host and completed **exit 0** with a real 768x1376 JPEG in 42.2 s, proving the host guard is scoped and does not short-circuit the working path. Three items recorded as NOT verified rather than omitted: driving the migrated frontend (still impossible, #639 stays open), the pt-BR recovery live (measured `html lang=pt` on `denon82` but exercised live only on `en-GB`), and the `en-GB`→`en` region reduction against a locale where region is load-bearing (`zh-Hans`/`zh-Hant`) — only two locales observed.

{gflow_cli-0.82.0 → gflow_cli-0.82.1}/docs/LIVE_VERIFICATION_v0.81.0.md

@@ -81,4 +81,8 @@ keyboard focus at launch.
 
 ## Post-tag evidence
 
-_Filled after the tag: PyPI publish, GitHub Release, back-merge._
+- **Release workflow:** [run 36786572400](https://github.com/ffroliva/gflow-cli/actions/runs/36786572400) — success (build-and-publish, mcp-registry publish).
+- **PyPI:** `gflow_cli-0.81.0-py3-none-any.whl` and `gflow_cli-0.81.0.tar.gz`, uploaded 2026-09-30T22:37Z.
+- **GitHub Release:** [v0.81.0](https://github.com/ffroliva/gflow-cli/releases/tag/v0.81.0), published 2026-09-30T22:37Z.
+- **Release PR:** #924, merged to `main` with a merge commit (`8fa5ea0c`).
+- **Back-merge:** `main` → `develop` (`be334758`).

{gflow_cli-0.82.0 → gflow_cli-0.82.1}/docs/LIVE_VERIFICATION_v0.82.0.md

@@ -85,4 +85,9 @@ true).
 
 ## Post-tag evidence
 
-_Filled after the tag: PyPI publish, GitHub Release, back-merge._
+- **Release workflow:** [run 36928643524](https://github.com/ffroliva/gflow-cli/actions/runs/36928643524) — success (build-and-publish, mcp-registry publish).
+- **PyPI:** `gflow_cli-0.82.0-py3-none-any.whl` and `gflow_cli-0.82.0.tar.gz`, uploaded 2026-10-01T21:26Z.
+- **GitHub Release:** [v0.82.0](https://github.com/ffroliva/gflow-cli/releases/tag/v0.82.0), published 2026-10-01T21:26Z.
+- **Release PR:** #937, merged to `main` with a merge commit (`00ff5720`).
+- **Back-merge:** `main` → `develop` (`e45532d0`).
+- **Install from PyPI:** `uvx --from gflow-cli==0.82.0 gflow --version` → `gflow, version 0.82.0`.

gflow_cli-0.82.1/docs/LIVE_VERIFICATION_v0.82.1.md

@@ -0,0 +1,91 @@
+# Live verification — v0.82.1
+
+> Evidence for the fixes in this patch release, run during development on 2026-10-01 and
+> 2026-10-02 (PRs #939, #940, #942). Where a check could not be run, the blocker is named.
+> Each row is a run that someone watched, not an inference.
+
+## Environment
+
+| | |
+|---|---|
+| Profile | `ci-probe` (real-browser Chrome strategy) |
+| Host Flow served | `flow.google.com` |
+| Transport | `ui_automation` → migrated composer |
+| OS | Windows 11 |
+| Veo credits spent | 1 × `veo-lite` clip (#898, consented); everything else $0 |
+
+## Summary
+
+| Change | Surface | Verified live? | Cost |
+|---|---|---|---|
+| A reCAPTCHA mint failure is a typed error with a measured retry flag (#915, PR #939) | token mint on the real page pool | ✅ BDD e2e `test_recaptcha_error_shape_bdd.py`: **2 passed** | $0 |
+| A busy catalog after a successful generation is a warning, not a failed run (#900, PR #940) | local SQLite | n/a: no Flow surface — verified on a real DB file with a real second connection | $0 |
+| A video's Flow workflow id is recorded (#898, PR #942) | `gflow video t2v` → catalog | ✅ one clip on flow.google.com | 1 veo-lite |
+| A repeated video start is a no-op (#898) | catalog | n/a: no Flow surface — offline | $0 |
+
+## #915 — reCAPTCHA mint failures
+
+Spike `scripts/dev/spike_recaptcha_error_shape.py`
+([findings](superpowers/spikes/2026-10-01-recaptcha-error-shape.md)), three runs, real page
+pool:
+
+- **Pool page at `about:blank`, client path:** `RecaptchaError` 3/3, identical → not retryable.
+- **flow.google.com root grid:** exit 36 before the mint (guarded).
+- **Project page, steady:** 5/5 and 3/3 mints OK.
+- **Mint racing a navigation:** "context destroyed" 3/3, every re-mint OK → retryable.
+- **Site-key read alone, 0–0.2 s into a navigation:** "context destroyed" at 0.02/0.05 s and
+  "no site key" at 0.1/0.2 s; every re-read OK. `document.readyState` is already `complete`
+  at 0.2–0.3 s with no key (Flow injects the script after load), so it cannot tell the two
+  apart; the page's scheme can (`about:` never gets the script). → a missing key is
+  retryable iff the page is `http(s)`.
+
+E2E (`@e2e @e2e_auth`, ci-probe): **2 passed** (30.9 s) — `about:blank` → typed,
+`retryable=False`, message names the page; a raced mint → typed, `retryable=True`, and the
+re-mint on the settled page succeeds.
+
+## #900 — busy catalog
+
+No Flow surface. The real dependency is SQLite, driven for real:
+`tests/data/test_store_transaction_errors.py` (lock past `busy_timeout`, nested BEGIN,
+`IntegrityError` still raw, non-sqlite error passes through with rollback) and
+`tests/cli/test_video_recording_busy.py` (a real recorder and DB file, the lock taken after
+the generation returns: exit 0, one `data.persistence_failed_after_success` warning with
+`detail` "database is locked", the STARTED row intact). The CLI test fails on the
+pre-fix `store.py` (exit 1, `error_unhandled OperationalError`) and passes with the fix.
+
+## #898 — the video workflow id
+
+`gflow video t2v --model veo-lite --profile ci-probe` (1 clip, consented):
+
+1. **Count:** 1 file, `7ec5466f-….mp4`.
+2. **Magic bytes:** `ftypisom` (MP4).
+3. **Shape:** 2,171,384 bytes.
+4. **Structlog:** `migrated.submit_observed` (`YhhmEf`) named workflow `dbb0737e…`; all 10
+   `migrated.status` replies matched it (the driver drops any that do not).
+5. **User-confirmable:** the catalog row has `flow_workflow_id = dbb0737e…` (NULL for every
+   video before the fix) and a lookup by that id returns the clip.
+
+The labs arm reads `media[0].workflowId`, present in every committed capture (02, 08, 09);
+it is covered offline against those captures.
+
+## Blockers (named, not waived)
+
+- **Accounts served labs:** labs answers 308 on the three profiles here that hold a live
+  Flow session, so the labs arms of #915 (where a user meets the retryable shapes) and #898
+  (`media[0].workflowId`) are unobserved live.
+
+## Pre-tag gates
+
+- **PRs** #939, #940, #942: CI 16/16 each, including SonarCloud.
+- **Councils:** #939 three reviewers (no must-fix; the site-key-read arm they asked for
+  falsified a first reading and changed the rule); #940 one reviewer (three should-fixes,
+  all fixed); #942 one reviewer (one must-fix — the labs workflow id was in the committed
+  captures — and two should-fixes, all fixed).
+- **`/gflow:doc-review`:** one auditor over completeness, cross-reference and drift — YELLOW,
+  no FAIL; both WARNs (MCP.md retry-override list, DATA_LAYER.md lock row) and the nits fixed
+  before the tag.
+- **`/gflow:check`:** all gates green on the release branch.
+
+## Post-tag evidence
+
+_Filled after the tag: PyPI publish, GitHub Release, back-merge._

{gflow_cli-0.82.0 → gflow_cli-0.82.1}/docs/MCP.md

@@ -147,7 +147,7 @@ agent-only exception in § 6.1, which overrides it to `false`), an unreachable
 UI arm (`UiModeUnavailableError`), and a partially-completed sync
 (`SyncPartialError`). That list is `errors.RETRYABLE_ERRORS`, but it is no longer
 the whole answer: `errors.is_retryable` consults the **instance** first, so a raise
-site can override its class. Two do today — Flow's `/about` redirect raises
+site can override its class. Among them, Flow's `/about` redirect raises
 `FlowAppError` with `retryable: false`, and since 2026-09-11 that is a **measurement**, not a
 preserved default: caught during a live occurrence, 5/5 consecutive attempts over ~3 minutes
 landed on `/about` again, on the account's own project with a healthy session — so a retry is
@@ -155,7 +155,11 @@ doomed and costs ~35 s each ([#756](https://github.com/ffroliva/gflow-cli/issues
 second is the migrated **agent-only composer**, which raises `FlowAgentUiError` with
 `retryable: false` because which composer an account gets is server-assigned and does not
 flap ([#799](https://github.com/ffroliva/gflow-cli/issues/799)). The
-*cause*, and whether it ever clears, remain unmeasured.
+*cause*, and whether it ever clears, remain unmeasured. The override also runs the other
+way: a reCAPTCHA mint failure (`type` `…/errors/recaptcha-mint`, exit 1, no request sent,
+no credit spent) is `retryable: true` when a navigation raced the mint or the page was a
+Flow page whose script had not loaded yet, and `false` on a page that is not a web page —
+each measured live ([#915](https://github.com/ffroliva/gflow-cli/issues/915)).
 Read the flag off the envelope; never re-derive it from the class list.
 Everything
 else (auth, content-policy, configuration, security) is terminal

{gflow_cli-0.82.0 → gflow_cli-0.82.1}/docs/PROJECT_STATUS.md

@@ -4,6 +4,30 @@
 
 ## Current release
 
+**v0.82.1 — alpha.** Three fixes for failures that misreported what happened.
+
+**A reCAPTCHA mint failure is a typed error (#915).** It was a bare `RuntimeError`: exit 1
+with no remediation, a hashed "Unknown Error" over MCP, and one failure ended a
+multi-prompt run past `--continue-on-error`. It now has its own `type`
+(`…/errors/recaptcha-mint`) and a `retryable` flag measured live: a mint that lost a race
+with a navigation, or ran before Flow injected its script, is retryable; a mint on a
+non-web page (`about:blank`) is not. Still exit 1 — branch on the `type`.
+
+**A busy catalog no longer fails a successful generation (#900).** A write blocked past the
+5 s lock timeout raised a raw sqlite error that missed every "recording failed after
+success" handler, so `gflow video` exited 1 for a clip that existed and was paid for.
+Catalog transactions now raise `DataStoreError`: the run succeeds with a warning.
+
+**A video's Flow workflow id is recorded (#898).** It was `NULL` for every clip, so the MCP
+task result's `flow_workflow_id` was always `null` for a video. A repeated start for a
+known media id is now a no-op instead of a crash.
+
+**Not verified here:** accounts served labs (labs answers 308 on the three profiles here
+that hold a live Flow session). Full ledger:
+[LIVE_VERIFICATION_v0.82.1](LIVE_VERIFICATION_v0.82.1.md).
+
+<details><summary>v0.82.0 — a run config builds a series from one image</summary>
+
 **v0.82.0 — alpha.** A `gflow run` config can build a series from one image, and every
 generation it makes is now tracked.
 
@@ -30,6 +54,8 @@ sign-in screen in the generation window (#925).
 **Not verified here:** accounts served labs (labs answers 308 on the three profiles here that hold a live Flow session). Full
 ledger: [LIVE_VERIFICATION_v0.82.0](LIVE_VERIFICATION_v0.82.0.md).
 
+</details>
+
 <details><summary>v0.81.0 — the generation browser opens off-screen</summary>
 
 **v0.81.0 — alpha.** The generation browser stops covering your desktop, a denied profile
@@ -356,6 +382,7 @@ migrated accounts (#795), and the agent-only composer driver (#799, #824 open).
 
 | Milestone | Status |
 |---|---|
+| A reCAPTCHA mint failure, a busy catalog and a missing video workflow id stop misreporting what happened (#915, #900, #898) | ✅ done (v0.82.1) |
 | A run config row generates from an earlier row's image or a local file, referenced in place with no re-upload (#913); `gflow run` successes recorded with lineage; refs were silently dropped since v0.52.0 (fixed, record corrected) | ✅ done (v0.82.0) |
 | The generation browser opens off-screen, placeable via `GFLOW_CLI_BROWSER_WINDOW_POSITION`, measured unchanged for image and video (#923); a write-denied profile raises `ProfileAccessError` instead of posing as lock contention (#919); `urllib3` CVE lock (#920) | ✅ done (v0.81.0) |
 | An unported image form is refused by name instead of crashing a `gflow run` batch with a misleading `RecaptchaError` — the browser transport stops pre-minting a token it never read (#891); a lost transfer is recorded as a generated clip with a visible `STATUS` (#896); `--resolution` and `nano2-lite` (#787); flow.google.com refusals read from the wire (#909); `auth login` stops on the `/about` identity re-check (#902); `oauthlib`/`pyjwt` CVE locks | ✅ done (v0.80.0) |

{gflow_cli-0.82.0 → gflow_cli-0.82.1}/docs/USAGE.md

@@ -2004,7 +2004,7 @@ shell scripts can branch on the failure mode without parsing stderr.
 | Code | Error class           | Meaning                                          | Remediation                                                |
 |------|-----------------------|--------------------------------------------------|------------------------------------------------------------|
 | `0`  | —                     | Success                                          | —                                                          |
-| `1`  | unhandled exception   | Anything not derived from `GFlowError` — **or a deliberate CLI verdict**: `gflow auth status` exits 1 for a dead/unverifiable session | Re-run with `--verbose`; for `auth status` follow the printed hint; file a bug if it persists |
+| `1`  | unhandled / unmapped  | Anything not derived from `GFlowError`, a typed error with no code of its own (e.g. a reCAPTCHA mint failure, `type` `…/errors/recaptcha-mint`, [#915](https://github.com/ffroliva/gflow-cli/issues/915) — branch on the `--json` `type`) — **or a deliberate CLI verdict**: `gflow auth status` exits 1 for a dead/unverifiable session | Re-run with `--verbose`; for `auth status` follow the printed hint; file a bug if it persists |
 | `2`  | usage error (Click)   | Bad usage / missing arg / profile missing        | Standard CLI usage error                                   |
 | `3`  | `AuthExpiredError`    | Session cookies rejected by Flow (401/403), or Flow served one of its OAuth/sign-in routes instead of the page gflow asked for ([#756](https://github.com/ffroliva/gflow-cli/issues/756)) | `gflow auth login --profile <name>` — **but read the error's own `remediation_hint` first.** `AisandboxAuthError` shares this code, and when `gflow credits` fails on an account migrated to `flow.google.com` re-logging in cannot help and can roll the profile's browser-strategy marker back ([#795](https://github.com/ffroliva/gflow-cli/issues/795), [#791](https://github.com/ffroliva/gflow-cli/issues/791)) |
 | `4`  | `RateLimitError`      | Quota / rate limit hit, exhausted retries        | Wait + reduce `GFLOW_CLI_CONCURRENCY`                      |
@@ -2048,6 +2048,12 @@ shell scripts can branch on the failure mode without parsing stderr.
 **Exit code 16 — data store / migration error.** Fires when:
 
 - The database file cannot be opened (filesystem permission or path issues).
+- A write could not take the database lock within 5 s (`database is locked` — another
+  gflow process was writing). After a **successful** `gflow image` or `gflow video`
+  generation this is only a warning (`data.persistence_failed_after_success`, naming the
+  media id): the file is saved and the run succeeds, it is just not recorded in the
+  catalog (#900). `gflow movie` logs its own `movie.persistence_failed_*` warning; a task
+  on the MCP worker queue records it as the task's error.
 - A migration fails or the migration checksum drifts from what the installed version expects.
 - The database has a **newer schema** than the installed gflow-cli (i.e. you downgraded after a migration already ran).
 

{gflow_cli-0.82.0 → gflow_cli-0.82.1}/docs/superpowers/memory/exit-code-16-data-store.md

@@ -11,7 +11,7 @@ Exit code 16 covers three error classes in `src/gflow_cli/errors.py`:
 
 All three are re-exported via `gflow_cli.exceptions`. Comment in `errors.py:423` documents the EXIT_CODE_MAP ordering invariant — see [[exit-code-map-ordering-invariant-test-pitfall]].
 
-**Contract:** exit code 16 fires ONLY for pre-Flow failures (DB open, migration check before any paid Flow call). Post-success persistence failures (generation completed, file downloaded, but `INSERT` failed) emit `data.persistence_failed_after_success` structlog event with `flow_media_id` + `local_path`, print a yellow console warning, and STILL return 0. This prevents scripts from retrying paid generations just because the local catalog couldn't be updated.
+**Contract:** exit code 16 fires for data-store failures that do NOT follow a successful paid generation: DB open, migration check before any paid Flow call, and (since #900) a catalog write blocked past `busy_timeout` outside a successful generation. Post-success persistence failures (generation completed, file downloaded, but `INSERT` failed) emit `data.persistence_failed_after_success` structlog event with `flow_media_id` + `local_path`, print a yellow console warning, and STILL return 0. This prevents scripts from retrying paid generations just because the local catalog couldn't be updated.
 
 **How to apply:**
 - User reports exit 16: check `GFLOW_CLI_DB_PATH` filesystem permissions, then check for "newer schema" — upgrade gflow-cli OR repoint `GFLOW_CLI_DB_PATH` to a compatible DB.

{gflow_cli-0.82.0 → gflow_cli-0.82.1}/docs/superpowers/memory/migrated-host-driver-wire-lessons.md

@@ -68,8 +68,9 @@ cost a real run to learn; each is now a unit test in `tests/api/transports/`.
 `transport.generate_images` / upscale / extend ever run, so transport-level
 `raise_if_migrated` guards never cover it. On a moved account that page is the
 `flow.google.com` grid (client-side handoff), which loads no `recaptcha/enterprise.js`,
-so `discover_site_key` raised `RecaptchaError` — a `RuntimeError` unmapped in
-`EXIT_CODE_MAP` — as exit 1 "unexpected" instead of exit 36. The guard now runs at the
+so `discover_site_key` raised `RecaptchaError` — then a `RuntimeError` unmapped in
+`EXIT_CODE_MAP` (a typed `GFlowError` since #915) — as exit 1 "unexpected" instead of
+exit 36. The guard now runs at the
 mint too (`client.py`, `at="mint_recaptcha_token"`); `git grep raise_if_migrated` is
 the current list of sites. Since #891 that mint guard covers only callers that really send
 the token (HTTP image transports, upscale, extend — #914); UI images no longer mint. Reviewing anything that adds a pre-transport step: ask

gflow_cli-0.82.1/docs/superpowers/spikes/2026-10-01-recaptcha-error-shape.md

@@ -0,0 +1,69 @@
+# Spike — what a reCAPTCHA mint failure looks like, and whether it is transient (#915)
+
+**Date:** 2026-10-01 · **Profile:** `ci-probe` (host Flow served: flow.google.com) ·
+**Cost:** $0 (mints only, tokens discarded) ·
+**Script:** [`scripts/dev/spike_recaptcha_error_shape.py`](../../../scripts/dev/spike_recaptcha_error_shape.py)
+
+## Question
+
+`RecaptchaError` subclasses `RuntimeError`, not `GFlowError` (#915), so it exits 1 with no
+Problem Details and aborts a batch past `--continue-on-error`. Making it a `GFlowError`
+needs one answer the code cannot give: **is it retryable?** A retry flag that invites a
+doomed loop is worse than none (`FlowHostMigratedError`, #639).
+
+## Where it can be raised (code, not live)
+
+`TokenMinter` (`api/recaptcha.py`) is the only raiser, reached only through
+`FlowApiClient._mint_recaptcha_token`, whose callers are:
+
+- `extend_video` (`gflow video extend`)
+- `_upsample_image_impl` (`gflow image upscale`)
+- image generation on a transport **without** `uses_page_owned_image_recaptcha` — the
+  experimental HTTP transports. The default `ui_automation` transport skips the client
+  mint since #891, so `gflow image t2i` / `image batch` / `gflow run` on the default
+  transport never raise it.
+
+No MCP tool exposes upscale or extend on `develop`; the worker reaches it only on an
+experimental transport.
+
+## Measurements (three runs on 2026-10-01, real page pool)
+
+| Arm | Path | Result |
+|---|---|---|
+| A. pool page at `about:blank` (#891/#914 state) | client `_mint_recaptcha_token` ×3 | `RecaptchaError` "Could not discover reCAPTCHA site key…" **3/3, identical** |
+| B. flow.google.com root grid | client path | `FlowHostMigratedError` (exit 36) — guarded before the mint |
+| C. flow.google.com `/project/<id>`, steady | `TokenMinter.mint` ×5 | **5/5 OK** (≈2.4k-char tokens) |
+| D. mint racing a navigation on that page | `TokenMinter.mint`, then re-mint after settle, ×3 | raced: `RecaptchaError` "evaluate failed … Execution context was destroyed" **3/3**; re-mint **3/3 OK** |
+| E. site-key **read** alone, 0–0.2 s into a navigation (council on #939: D's failures were all the *execute* step) | `discover_site_key`, then re-read after settle | 0 s: OK ×2 · 0.02/0.05 s: "context destroyed" ×2 · **0.1/0.2 s: "no site key" ×2** · every re-read **OK** |
+| F. why E's 0.1–0.2 s rows said "no key" | `document.readyState` + key at the failure point | 0.1/0.15 s `interactive`, no key · **0.2/0.3 s `complete`, no key** · 0.5 s `complete`, key present · `about:blank` `complete`, no key |
+
+## Reading
+
+- **An execute failure, or a failed site-key read, is a page-state race that clears**
+  (D, E). The same page mints once it settles. → `retryable=True`.
+- **A missing site key is NOT always deterministic.** The first reading (from A alone)
+  said it was; E falsified that: 0.1–0.2 s into a load a Flow page has no key yet, and a
+  re-read finds it. **F falsified the obvious discriminator too:** `document.readyState`
+  is already `complete` at 0.2–0.3 s with no key — Flow injects the script after load —
+  so "still loading" cannot be read off the document.
+- **What does separate the two is the page itself.** `about:blank` is not a web page and
+  never gets the script (A: 3/3); a Flow `https` page mid-load does (E: 4/4 re-reads).
+  → missing key is `retryable` **iff** the page URL is `http(s)`. The one `https` page
+  known to lack the script for good, the flow.google.com root grid, is refused before the
+  mint (B: exit 36), so it never reaches this flag.
+- A healthy project page does not fail (C), so there is no evidence of a background
+  flake rate that would make the whole class retryable.
+- **Where a user meets it.** Through the client, any flow.google.com page is refused
+  before the mint (B), so the retryable shapes reach a user only on a page Flow served
+  from labs. A navigation destroying a page's context is Playwright behaviour, not a
+  host's, so the flag is measured where it can be; the labs arm itself is unobserved.
+
+## Not measured
+
+- An empty token returned by `grecaptcha` (the third raise site): not produced here.
+  It keeps the class default (not retryable) until it is observed.
+- An `https` page that finished loading and will never get the script (a labs layout
+  change): would be flagged retryable and cost one wasted re-run — no request is sent and
+  no credit spent by a failed mint. Not observed.
+- Labs-served accounts: none reachable (labs answers 308 on the three profiles here
+  that hold a live Flow session).

{gflow_cli-0.82.0 → gflow_cli-0.82.1}/plugins/gflow/.claude-plugin/plugin.json

@@ -1,7 +1,7 @@
 {
   "name": "gflow",
   "displayName": "gflow-cli",
-  "version": "0.82.0",
+  "version": "0.82.1",
   "description": "Drive Google Flow from your agent: Veo text-to-video, image-to-video and clip extension, Imagen images, reusable Characters and scene composition.",
   "author": {
     "name": "Flavio Oliva",