gflow-cli 0.54.0__tar.gz → 0.56.0__tar.gz

{gflow_cli-0.54.0 → gflow_cli-0.56.0}/.codex-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "gflow",
-  "version": "0.54.0",
+  "version": "0.56.0",
   "description": "Reusable development and operations workflows for gflow-cli",
   "author": {
     "name": "Flavio Oliva",

{gflow_cli-0.54.0 → gflow_cli-0.56.0}/.env.template

@@ -133,6 +133,16 @@
 # changes; don't exceed 8 without measuring. Shipped in v0.4.0a2.
 # GFLOW_CLI_CONCURRENCY=1
 
+# Seconds to wait for another gflow process to release the profile lease
+# before giving up with ProfileLockedError (0 = fail fast, the default).
+# Holders always run to completion; the waiter takes over on release (#478).
+# GFLOW_CLI_LEASE_WAIT_SECONDS=0
+
+# Once-a-day best-effort PyPI check printing a one-line notice when a newer
+# gflow-cli exists (#479). Never blocks; skipped in CI and for
+# editable/source installs. Set to 0 to disable.
+# GFLOW_CLI_UPDATE_CHECK=1
+
 # Run Playwright in headless mode for non-`auth login` commands (true|false).
 # Default: false — headed real Chrome is the PRODUCTION default, not a
 # fallback. The ui_automation transport (the only production transport)

{gflow_cli-0.54.0 → gflow_cli-0.56.0}/AGENTS.md

@@ -8,15 +8,15 @@ Supported tools that auto-discover this file: Cursor, Codex, Aider, Antigravity
 
 - Unofficial Python CLI for [Google Flow](https://labs.google/fx/tools/flow) — drives Veo (image-to-video, text-to-video) and Imagen (text-to-image) generations from the terminal by reverse-engineering Flow's private REST API at `aisandbox-pa.googleapis.com`.
 - Python 3.11+ · `uv`-managed · `hatchling` builds · Playwright Chromium transport · `pyright` strict · `ruff` · `pytest`.
-- Single-package modular monolith. Top-level modules under `src/gflow_cli/`: `api/`, `auth/`, `data/`, `mcp/`, `services/`, `tools/`, `ui/`, `worker/`, `browser_manager.py`, `cli.py`, `_cli_helpers.py`, `cli_character.py`, `cli_data.py`, `cli_image.py`, `cli_instructions.py`, `cli_models.py`, `cli_movie.py`, `cli_run.py`, `cli_scene.py`, `cli_tools.py`, `cli_video.py`, `chain.py`, `chain_manifest.py`, `composition.py`, `config.py`, `errors.py`, `exceptions.py`, `image_batch.py`, `movie_manifest.py`, `observability.py`, `paths.py`, `profile_store.py`.
+- Single-package modular monolith. Top-level modules under `src/gflow_cli/`: `api/`, `auth/`, `data/`, `mcp/`, `services/`, `tools/`, `ui/`, `worker/`, `browser_manager.py`, `cli.py`, `_cli_helpers.py`, `diagnostics.py`, `json_output.py`, `media.py`, `profile_lease.py`, `redaction.py`, `storage.py`, `winsec.py`, `cli_project.py`, `cli_character.py`, `cli_data.py`, `cli_image.py`, `cli_instructions.py`, `cli_models.py`, `cli_movie.py`, `cli_run.py`, `cli_scene.py`, `cli_tools.py`, `cli_video.py`, `chain.py`, `chain_manifest.py`, `composition.py`, `config.py`, `errors.py`, `exceptions.py`, `image_batch.py`, `movie_manifest.py`, `observability.py`, `paths.py`, `profile_store.py`.
 - Command surface: `gflow auth`, `gflow image` (t2i/i2i/batch/upload/upscale), `gflow video` (t2v/i2v/r2v/chain — no `batch` subcommand; the nonfunctional stub was removed, loop `gflow video t2v`/`i2v` from the shell for multi-clip runs), `gflow character` (create/list/show/rm/voices — reusable project-scoped Flow Character entities), `gflow scene` (create/show — Add Clip / Scenes, with `create --output` for credit-free server-side extended video), `gflow instructions` (persistent Agent-Mode brief cards — add/list/enable/disable/rm/apply/toggle-mode, credits-free, `--project` required), `gflow movie` (run/template — multi-scene manifest pipeline), `gflow tools` (list/show/run — prompt-rewriting tools, also `--tool` on generation commands), `gflow data` (catalog queries), `gflow models`, `gflow run`, `gflow mcp` (run/setup — stdio MCP server), and `gflow serve` (Streamable HTTP at `/mcp`; `--transport sse` is deprecated).
-- Requires a Google AI Ultra or Pro subscription with Flow access. All generations bill against the user's own Google account.
+- Works with any Google account that has Flow access. All generations bill against the user's own Google account.
 
 ## Headed-browser dependency (architectural reality)
 
 gflow-cli currently drives Flow via a **real Chrome session managed by Playwright** — `ui_automation` transport. Google's auth + reCAPTCHA stack rejects Playwright's bundled Chromium and most headless approaches. This is the project's defining trade-off:
 
-- ✅ Works end-to-end against live Pro/Ultra accounts.
+- ✅ Works end-to-end against live Google accounts.
 - ❌ Requires a saved Chrome profile, a display server for one-time login, and ~150 MB for Chromium.
 - ❌ Cannot run on serverless / headless CI workers without prerecorded profile transplant.
 - ❌ Per-account horizontal concurrency is capped by what one warm Page pool can drive.
@@ -58,7 +58,7 @@ Or invoke the wrapper: `/gflow:check`.
 
 - Type hints everywhere; `pyright` strict on `src/gflow_cli`.
 - Structured logging only (`structlog`) — **never** raw `print()` or `import logging` in `src/`.
-- Errors as RFC 9457 Problem Details with stable per-class exit codes (3–30, e.g. 11 is `ConfigurationError` — including `ProfileLockedError` for same-profile lease contention, 16 is the `DataStoreError` family, 19 `SceneConcatError`, 20 `FrameExtractionError`, 21 `ChainPartialError`, 22 `UpscaleUnavailableError`, 25 `FlowAgentUiError`, 28 `UiModeUnavailableError`, 29 `MentionIndexUnavailableError`, 30 `QueueSchemaError`). See `src/gflow_cli/errors.py::EXIT_CODE_MAP` for the complete mapping.
+- Errors as RFC 9457 Problem Details with stable per-class exit codes (3–31, e.g. 11 is `ConfigurationError` — including `ProfileLockedError` for same-profile lease contention, 16 is the `DataStoreError` family, 19 `SceneConcatError`, 20 `FrameExtractionError`, 21 `ChainPartialError`, 22 `UpscaleUnavailableError`, 25 `FlowAgentUiError`, 28 `UiModeUnavailableError`, 29 `MentionIndexUnavailableError`, 30 `QueueSchemaError`). See `src/gflow_cli/errors.py::EXIT_CODE_MAP` for the complete mapping.
 - 100-char line length, `ruff` configured. Imports sorted by `ruff` (isort rules).
 - **YAGNI / least-code**: prefer the smallest change that works. No speculative abstractions (interface/factory with one implementation), no config or flags nobody sets, no dead constants/helpers, no reinventing the stdlib. Review carries this as its own lens — the **D14 over-engineering** dimension of [`pr-council-review`](skills/pr-council-review/SKILL.md) (baseline, always runs). Its rubric is portable; the `ponytail` plugin (see CONTRIBUTING) is an optional accelerant, not a dependency.
 - **MCP & CLI Schema Symmetry**: Any updates or additions to user-facing CLI command parameters (e.g., `gflow image t2i`, `gflow video`) must be mirrored in the corresponding MCP tool definitions. Never add option/argument fields to Click commands without updating the MCP server implementation. This symmetry is enforced programmatically in CI via `tests/mcp/test_cli_parity.py` (every CLI leaf command needs a mapped MCP tool or an explicit, reasoned exemption) plus the schema checks in `tests/mcp/test_server.py`.

{gflow_cli-0.54.0 → gflow_cli-0.56.0}/CHANGELOG.md

@@ -7,6 +7,57 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [0.56.0] — 2026-08-13
+
+### Added
+
+- **Once-a-day PyPI update notice (#479).** gflow now prints a one-line stderr notice when a newer version is on PyPI: cache-served (zero added latency — a stale cache refreshes on a background daemon thread for the next run), capped at one poll per day even when the poll fails, and never blocks or fails a command. Skipped in CI (`CI` env var), for editable/local-source installs (PEP 610 detection), and when `GFLOW_CLI_UPDATE_CHECK=0`.
+
+- **Opt-in bounded wait for profile-lease contention (#478).** `GFLOW_CLI_LEASE_WAIT_SECONDS=N` makes a command that hits same-profile contention poll the kernel lock (0.5 s cadence, sync and async call sites alike) and take over as soon as the current holder — a CLI command or a `gflow serve` daemon queue task, both of which release at their natural end — finishes; on timeout it raises the same `ProfileLockedError` (exit 11). Default `0` keeps the historical fail-fast. Triage note: the issue's fuller cooperative-handoff protocol (release-request channel, minimum-hold window) was validated against the actual architecture and dropped — the daemon acquires the lease per task, so every holder's safe release point already coincides with lease release; holders are never asked to release early, which satisfies the no-release-mid-call requirement by construction. Same-process contention always fails fast (waiting on yourself deadlocks).
+
+- **Chromium downgrade guard for persisted profiles (#477).** Opening a profile with an older Chromium major version than last wrote it triggers Chromium's downgrade cleanup, which can shred the newer session store and surface as a mystery post-upgrade logout. Every bundled-Chromium open of a persisted profile (generation client, UI-automation transport, the experimental transports, headless verification probe) now compares the profile's `Last Version` against the active engine's bundled Chromium (`playwright`, or `patchright` when selected) and refuses on a major-version downgrade (`ProfileEngineDowngradeError`, exit 11; the fail-closed verification probe reports it as a verification failure) with an error naming both versions and the remedy. Best-effort: `chrome`-strategy profiles and unknown/unparseable versions skip the check; same-major build rollbacks are allowed; `gflow auth login` stays unguarded as the recovery path.
+
+### Fixed
+
+- **Mode-switch drift errors now name the right evidence (#493).** An external
+  report showed a third, unrecognized Flow editor layout (composer frame slots +
+  Agent toggle, no classic `crop_*` settings button) falling through to
+  `UiSelectorDriftError` (exit 23). The fall-through detail now states that no
+  known Flow cohort matched — i.e. the editor may be a new layout this version
+  does not recognize — and the class remediation no longer asks for a "debug
+  screenshot from this message" that the mode-switch probe never writes: it
+  points at the artifacts that actually exist (the PII-safe
+  `diag_mode_switch_miss.json` DOM signature and/or the referenced screenshot,
+  plus the incident bundle's `report.md`). Recognizing the new variant itself is
+  tracked in #493 and needs an affected account's diagnostics JSON. A post-merge
+  `/code-review` pass extended the same correction to the neighboring surfaces it
+  had missed: `FlowAgentUiError`'s privacy caveat now also covers the standalone
+  `debug_forced_agent_ui.png` its raise site writes, and the KNOWN_ISSUES
+  fallback points at the incident bundle's `ui.json` (which carries the DOM
+  signature) instead of `report.md` (which does not).
+
+## [0.55.0] — 2026-08-13
+
+### Added
+
+- **`gflow mcp setup` is implemented (#475).** The long-stubbed command now writes the gflow MCP server entry into the target client's config: `--target claude-desktop` (default), `cursor`, or `vscode` (VS Code's `servers` + `"type": "stdio"` schema). Non-destructive by construction: existing config content is merged, a pre-existing file is backed up as `<name>.gflow-backup`, a manual `gflow-cli` entry is converged in place instead of duplicated, and a corrupt config fails loud (exit 11) without ever being touched. Also fixed the bogus `%APPDATA%\Castano\Claude` path in docs/MCP.md.
+- **Incident bundles now include a pre-filled bug-report template (#476).** On every captured incident, the recorder stages `report.md` alongside the existing artifacts: version/platform, error class + exit code + retryability, phase/route, and pointers to the captured evidence — built exclusively from the allowlisted manifest fields (never raw exception text). The CLI error message now prints the report path next to the bundle path, and retention treats `report.md` as recorder-owned. Downgrade caveat: gflow <= 0.54.x does not recognize `report.md`, so its retention classifies new bundles as unknown and stops pruning them — delete `<GFLOW_CLI_HOME>/incidents` manually if you downgrade. The report is meant to be copied out of the bundle before editing (the template says so), and the e2e privacy scanner now leak-scans `*.md` alongside the JSON artifacts.
+
+### Changed
+
+- **`gflow auth status` now proves the Flow session and exits 0/1 (#471).** The command previously only checked that the profile directory and cookies file exist — it could report OK on a dead session. It now runs the fast `verify_flow_profile` probe (cookie snapshot + Flow session endpoint; no browser, no credits) and exits 0 only on a verified session, printing the verified account email; any other outcome exits 1 with a `gflow auth login` remediation hint. Fail-closed: an unreachable endpoint is a failure, never an OK.
+
+### Fixed
+
+- **Removed the false "requires a Google AI Ultra or Pro subscription" claim across all docs.** Any Google account with Flow access can use gflow-cli — a paid plan only affects credit allowances and tier-gated features (e.g. 4K upscale stays Ultra-only). Swept README, AGENTS.md, DISCLAIMER.md, USER_GUIDE, CONFIGURATION, AUTHENTICATION, DEBUGGING, the medium tutorial, and the gflow-cli skill; factual tier-gating notes are unchanged.
+
+### Security
+
+- **MCP tools no longer leak raw exception text to clients (#473).** Every registered MCP tool now routes through one error funnel: `GFlowError`s keep their structured RFC-9457 problem-details envelope (with the shared `retryable` flag), while unexpected exceptions return a masked envelope carrying only the exception class name — messages can embed filesystem paths, profile names, or token material, so the full text goes exclusively to the server-side structured log (`mcp.tool.unexpected_error`). Previously only 3 of 11 tools had any funnel, and those leaked `str(exc)` as the `detail` field; the rest leaked through the framework's default error path. Guarded by a registry-introspection test that fails if any future tool bypasses the funnel.
+
+- **Windows profile dirs are now really restricted to the current user (#472).** POSIX `chmod 0700`/`0600` is a no-op on Windows, so a profile created under a custom `GFLOW_CLI_HOME` on a shared/world-readable volume inherited that visibility — with the live Google session cookies inside. `gflow auth login` now applies an explicit `icacls` DACL before the browser runs: inheritance stripped, a single owner-only ACE (by SID, locale-safe), children reset to inherit it. Two-step sequence verified empirically (the naive `/t` grant leaves files without effective access). Existing profiles created before this release are swept on their next use: the client's browser-launch path runs a marker-gated hardening pass once per profile. Traversal profile names are rejected before any directory is created or re-ACLed. Best-effort — an ACL failure is logged (with icacls exit code and stderr) and never blocks login; no-op off Windows.
+- **Secret settings can no longer leak through a `Settings` dump (#474).** `llm_api_key` (`GFLOW_CLI_LLM_API_KEY`) and `daemon_token` (`GFLOW_CLI_DAEMON_TOKEN`/`GFLOW_DAEMON_TOKEN`) are now stored as Pydantic `SecretStr`, so `repr()`, `str()`, and `model_dump_json()` of the settings object mask them by construction — defense-in-depth on top of the existing logging-boundary redaction. Values are unwrapped only at the single point of use (the prompt-tools LLM client).
+
 ## [0.54.0] — 2026-08-12
 
 ### Changed
@@ -2774,7 +2825,9 @@ shell-script template that branches on these codes.
 
 First skeleton. Not functional end-to-end yet.
 
-[Unreleased]: https://github.com/ffroliva/gflow-cli/compare/v0.54.0...HEAD
+[Unreleased]: https://github.com/ffroliva/gflow-cli/compare/v0.56.0...HEAD
+[0.56.0]: https://github.com/ffroliva/gflow-cli/compare/v0.55.0...v0.56.0
+[0.55.0]: https://github.com/ffroliva/gflow-cli/compare/v0.54.0...v0.55.0
 [0.54.0]: https://github.com/ffroliva/gflow-cli/compare/v0.53.1...v0.54.0
 [0.53.1]: https://github.com/ffroliva/gflow-cli/compare/v0.53.0...v0.53.1
 [0.53.0]: https://github.com/ffroliva/gflow-cli/compare/v0.52.0...v0.53.0

{gflow_cli-0.54.0 → gflow_cli-0.56.0}/CONFIGURATION.md

@@ -7,7 +7,9 @@
 | `GFLOW_CLI_HOME` | Platform data dir | Root directory for profiles and config. |
 | `GFLOW_CLI_OUTPUT_DIR` | `~/Downloads/gflow-cli` | Where generated images and videos are saved. |
 | `CHROME_BINARY` | (autodetect) | Override Chrome binary path. Falls back to platform-standard locations. |
-| `GFLOW_CLI_CONCURRENCY` | `4` | Maximum parallel API requests per batch. |
+| `GFLOW_CLI_CONCURRENCY` | `1` | Page-pool size per batch (raising above 1 currently has no effect — see [docs/CONFIGURATION.md](docs/CONFIGURATION.md)). |
+| `GFLOW_CLI_UPDATE_CHECK` | `1` | Once-a-day stderr notice when a newer gflow-cli is on PyPI; `0` disables (v0.56.0). |
+| `GFLOW_CLI_LEASE_WAIT_SECONDS` | `0` | Bounded wait on same-profile lease contention instead of fail-fast (v0.56.0). |
 | `GFLOW_CLI_LLM_BASE_URL` | Google compat endpoint | OpenAI-compatible endpoint for the prompt tools (`--tool`). The on/off switch. |
 | `GFLOW_CLI_LLM_API_KEY` | _(unset)_ | Credential for that endpoint. Optional — omit for a keyless local gateway. [Google key](https://aistudio.google.com/apikey). |
 | `GFLOW_CLI_LLM_MODEL` | _(unset)_ | Model, and therefore the provider selector. Unset = the gateway chooses. |

{gflow_cli-0.54.0 → gflow_cli-0.56.0}/DISCLAIMER.md

@@ -19,7 +19,7 @@ If you need a stable, supported, contractual API, use the [official Google Gen A
 
 To use `gflow-cli` you must:
 
-1. **Own a valid Google account** with active access to [Google Flow](https://labs.google/fx/tools/flow) — typically an [AI Ultra or AI Pro](https://gemini.google/subscriptions/) subscription as of late 2025.
+1. **Own a valid Google account** with active access to [Google Flow](https://labs.google/fx/tools/flow) — any Google account works; a paid plan ([AI Pro or AI Ultra](https://gemini.google/subscriptions/)) only affects credit allowances and tier-gated features.
 2. **Comply with Google's terms of service** for that account, including:
    - [Google Terms of Service](https://policies.google.com/terms)
    - [Generative AI Additional Terms](https://policies.google.com/terms/generative-ai)
@@ -36,10 +36,10 @@ The maintainer asks that you do **not** use `gflow-cli` to:
 
 - Generate content that violates the [Generative AI Prohibited Use Policy](https://policies.google.com/terms/generative-ai/use-policy) (CSAM, harassment, IP infringement, etc.).
 - Circumvent Google's billing, quota, or rate limits via account-rotation farms.
-- Automate Flow access for users who do not have their own Ultra/Pro entitlement.
+- Automate Flow access for users who do not have their own Flow access entitlement.
 - Re-sell access to Veo via this tool as a "wrapped" commercial service.
 
-These uses can get **your** Google account banned, can put the maintainer in a difficult position with Google, and are antithetical to why this project exists (which is to let legitimate Ultra/Pro subscribers use credits they already paid for, more efficiently).
+These uses can get **your** Google account banned, can put the maintainer in a difficult position with Google, and are antithetical to why this project exists (which is to let legitimate Flow user subscribers use credits they already paid for, more efficiently).
 
 ## Takedown policy
 

{gflow_cli-0.54.0 → gflow_cli-0.56.0}/KNOWN_ISSUES.md

@@ -181,6 +181,32 @@ accepted. Two corrections fixed it (verified end-to-end on macOS Apple Silicon):
 
 Evidence: [LIVE_VERIFICATION_v0.23.0](docs/LIVE_VERIFICATION_v0.23.0.md).
 
+### New Flow editor variant (composer frame slots + Agent toggle) is not recognized
+
+- **Status:** **Open** — awaiting a DOM signature from an affected account;
+  tracked in [#493](https://github.com/ffroliva/gflow-cli/issues/493)
+- **Severity:** High · **Affects:** `gflow image` / `gflow video` generation on
+  accounts that received the new editor, any locale
+
+A third Flow editor layout was reported on 2026-08-13 (macOS, v0.55.0; also
+v0.53.1): frame-slot buttons ("Initial"/"Final") sit directly on the composer
+next to an "Agent" pill, and there is **no** classic `crop_*` aspect/settings
+button for gflow to drive. The layout matches **none** of the known cohort
+indicators (agentic chat composer, full-page media library), so the mode-switch
+step falls through to `UiSelectorDriftError` (**exit 23**) instead of the
+retryable exit-25 cohort classification. The selector cascade itself is
+locale-invariant — the reporter's Portuguese UI is unrelated.
+
+The error detail now names this hypothesis and the run writes a PII-safe DOM
+signature to `diag_mode_switch_miss.json` when it can be captured (structural
+allowlist: ligature names, tag counts — no cookies, tokens, prompts, or page
+text; if the file is absent, attach the incident bundle instead — its
+`ui.json` carries the same DOM signature). **If you hit this, attach that JSON file to
+[#493](https://github.com/ffroliva/gflow-cli/issues/493)** — recognizing the
+variant (and restoring a clean retryable classification, or full support) needs
+exactly that DOM signature. Re-running later may land the classic editor if
+the rollout is still an A/B arm.
+
 ### Flow's new full-page media-library UI breaks entity attach (A/B rollout)
 
 - **Status:** **Open** — Flow-side staged rollout; tracked in
@@ -203,10 +229,10 @@ classic `crop_*` aspect/mode control, so `gflow image`/`gflow video` can't drive
 generation. Rather than the old opaque `UiSelectorDriftError` "file a bug", the
 mode-switch raise site now runs a runtime DOM scan and raises a clear, **retryable**
 `FlowAgentUiError` (**exit 25**, "this cohort flaps; retry shortly"), and dumps a
-DOM-signature diagnostics artifact (`diag_mode_switch_miss.json` + a full-page
-screenshot) for reporting. The cohort is server-assigned per page load and flaps
-within ~12h, so a re-run often lands the classic UI. Driving the new UI directly
-is still out of scope.
+DOM-signature diagnostics artifact (`diag_mode_switch_miss.json`; the incident
+bundle carries a screenshot under `sensitive/`) for reporting. The cohort is
+server-assigned per page load and flaps within ~12h, so a re-run often lands
+the classic UI. Driving the new UI directly is still out of scope.
 
 **How to tell which UI your account has:** in the Flow web editor, click
 **Add Media** — a small dialog means the old (working) UI; a navigation to a
@@ -846,7 +872,7 @@ key — surfaces as a `RuntimeError` that `auth/cookies.py` normalizes to
 
 - **Status:** Mitigated (crash → typed fail-fast rejection) · **Severity:** Low · **Affects:** all versions
 
-Chromium refuses to open two persistent contexts on the same `user-data-dir` simultaneously. Historically this surfaced as an unhelpful Chromium "ProcessSingleton: profile is locked" error partway through a run. As of the profile-lease hardening (production-readiness plan, slice D1/D3), gflow-cli enforces this itself: a cross-process advisory lock (`ProfileLease`, kernel `flock` on POSIX / `msvcrt.locking` on Windows) guards every profile directory. A second `gflow` invocation, `gflow serve` daemon task, or MCP call against an already-leased profile is rejected **immediately** — before any Chrome process starts — with a typed `ProfileLockedError` (**exit code 11**); it never waits and never silently corrupts the profile.
+Chromium refuses to open two persistent contexts on the same `user-data-dir` simultaneously. Historically this surfaced as an unhelpful Chromium "ProcessSingleton: profile is locked" error partway through a run. As of the profile-lease hardening (production-readiness plan, slice D1/D3), gflow-cli enforces this itself: a cross-process advisory lock (`ProfileLease`, kernel `flock` on POSIX / `msvcrt.locking` on Windows) guards every profile directory. A second `gflow` invocation, `gflow serve` daemon task, or MCP call against an already-leased profile is rejected **immediately** by default — before any Chrome process starts — with a typed `ProfileLockedError` (**exit code 11**); it never silently corrupts the profile. Since #478, setting [`GFLOW_CLI_LEASE_WAIT_SECONDS`](docs/CONFIGURATION.md#gflow_cli_lease_wait_seconds) opts a waiter into a bounded wait that takes over as soon as the current holder finishes (holders always run to completion and are never asked to release early; same-process contention still fails fast — waiting on yourself would deadlock).
 
 **Workaround:** use different profiles for parallel work — different profiles acquire independent leases and run fully concurrently.
 
@@ -858,7 +884,7 @@ gflow image batch ./batch-a.tsv --profile work
 gflow image batch ./batch-b.tsv --profile personal
 ```
 
-**Roadmap:** Phase 4 (v0.4.0a2) added a per-worker Page pool on one shared BrowserContext (`GFLOW_CLI_CONCURRENCY=N`), intended to let one `gflow-cli` process fan out multiple in-flight generations, but no current CLI command drives more than one generation at a time through it — the only feature that did (a manifest-driven video batch runner) never worked and was removed. Cross-process same-profile serialization remains a hard constraint (Chromium can only own one persistent context per `user-data-dir`) — but it is now a clean, typed, fail-fast rejection instead of an unstructured crash. Multiple shells against the same profile remains a "use different profiles" workaround; there is no queueing/waiting mode.
+**Roadmap:** Phase 4 (v0.4.0a2) added a per-worker Page pool on one shared BrowserContext (`GFLOW_CLI_CONCURRENCY=N`), intended to let one `gflow-cli` process fan out multiple in-flight generations, but no current CLI command drives more than one generation at a time through it — the only feature that did (a manifest-driven video batch runner) never worked and was removed. Cross-process same-profile serialization remains a hard constraint (Chromium can only own one persistent context per `user-data-dir`) — but it is now a clean, typed, fail-fast rejection instead of an unstructured crash. Multiple shells against the same profile can either use different profiles (fully parallel) or opt into the bounded wait via `GFLOW_CLI_LEASE_WAIT_SECONDS` (serialized, waiter takes over when the holder finishes).
 
 ---
 

{gflow_cli-0.54.0 → gflow_cli-0.56.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: gflow-cli
-Version: 0.54.0
+Version: 0.56.0
 Summary: Unofficial CLI for Google Flow — drive Veo image-to-video generations from the terminal.
 Project-URL: Homepage, https://github.com/ffroliva/gflow-cli
 Project-URL: Issues, https://github.com/ffroliva/gflow-cli/issues
@@ -84,7 +84,7 @@ Description-Content-Type: text/markdown
 [![Tests: TDD](https://img.shields.io/badge/tests-TDD-brightgreen.svg)](CONTRIBUTING.md)
 [![Quality Gate Status](https://sonarcloud.io/api/project_badges/measure?project=ffroliva_gflow-cli&metric=alert_status)](https://sonarcloud.io/summary/new_code?id=ffroliva_gflow-cli)
 
-> ⚠️ **Read this before you install.** gflow-cli is **unofficial, alpha, and reverse-engineered — not affiliated with Google**. It drives a headed browser on *your own* Google Flow session, so treat it as your own account risk: automation is subject to Google's ToS, and endpoints or UI can change without notice. It requires a Google AI **Ultra or Pro** subscription with Flow access, and every generation bills your account. Read the full [DISCLAIMER](DISCLAIMER.md).
+> ⚠️ **Read this before you install.** gflow-cli is **unofficial, alpha, and reverse-engineered — not affiliated with Google**. It drives a headed browser on *your own* Google Flow session, so treat it as your own account risk: automation is subject to Google's ToS, and endpoints or UI can change without notice. It works with **any Google account** that has Flow access, and every generation bills against your account's Flow credit allowance. Read the full [DISCLAIMER](DISCLAIMER.md).
 >
 > 💳 **What failure costs you.** Credits are only spent on Veo *video* generation — images and composition ops are free, so most breakage costs nothing. When Flow's UI drifts mid-run, the CLI fails fast and loudly with distinct exit codes (e.g. selector drift = exit 23) instead of resubmitting, and batch items are recorded locally *before* submission so a broken run never silently burns credits on a stale state. See [KNOWN_ISSUES](KNOWN_ISSUES.md) for the current risk list.
 >
@@ -92,7 +92,7 @@ Description-Content-Type: text/markdown
 
 ## Why gflow-cli?
 
-You pay for Google AI Ultra or Pro, you have Veo credits, and you run real batch work. gflow-cli gives you:
+You have a Google account with Flow access, you have Veo credits, and you run real batch work. gflow-cli gives you:
 
 - **Batch generation.** Loop prompts straight from the shell: `for p in $(cat prompts.txt); do gflow image t2i "$p"; done`. Image batching plus `gflow video t2v` / `i2v` / `r2v` all ship today.
 - **Consistent subjects.** `gflow character create` mints a Flow Character (face and body reference) so the same person appears from one generation to the next.
@@ -100,7 +100,7 @@ You pay for Google AI Ultra or Pro, you have Veo credits, and you run real batch
 - **Pipelines.** Wire Veo into your content automation, AI-video stack, or batch experiments.
 - **Terminal-native.** After one `gflow auth login`, you stay in the shell. No clicking through dialogs.
 
-Same Veo and Imagen models, same quality, same Ultra/Pro billing, now programmatic.
+Same Veo and Imagen models, same quality, same billing against your own Google account, now programmatic.
 
 ## 60-second quick start
 
@@ -144,7 +144,7 @@ One command in, real Flow output back. Left: `gflow image t2i` generating a phot
 
 ![gflow image t2i runs a single 9:16 prompt, streams structlog output, and writes a PNG to disk](https://raw.githubusercontent.com/ffroliva/gflow-cli/main/docs/assets/example-run.gif)
 
-A single `gflow image t2i "..." --aspect 9:16 --model nano2` call against a logged-in Pro/Ultra profile. The terminal streams the run's `structlog` JSON, then lists the written PNG. Chromium drives the Flow editor silently in the background.
+A single `gflow image t2i "..." --aspect 9:16 --model nano2` call against a logged-in Flow profile. The terminal streams the run's `structlog` JSON, then lists the written PNG. Chromium drives the Flow editor silently in the background.
 
 Reproduce the recording with [`scripts/record_demo.ps1`](scripts/record_demo.ps1) (Windows, OBS, ffmpeg, gifski). More formats, including the side-by-side split-screen: **[docs/DEMOS.md](docs/DEMOS.md)**.
 
@@ -202,7 +202,7 @@ Full milestone history lives in [CHANGELOG.md](CHANGELOG.md). Where the project
 
 ## License & legal
 
-[MIT License](LICENSE) © 2026 Flavio Oliva (`ffroliva`). The MIT license covers `gflow-cli`'s code only. It grants no rights to Flow, Veo model output, or any Google service. Google's own terms (Labs Additional Terms, Ultra/Pro subscription terms) govern your generations. See the [DISCLAIMER](DISCLAIMER.md).
+[MIT License](LICENSE) © 2026 Flavio Oliva (`ffroliva`). The MIT license covers `gflow-cli`'s code only. It grants no rights to Flow, Veo model output, or any Google service. Google's own terms (Labs Additional Terms and any plan-specific subscription terms) govern your generations. See the [DISCLAIMER](DISCLAIMER.md).
 
 ## Acknowledgements
 

{gflow_cli-0.54.0 → gflow_cli-0.56.0}/README.md

@@ -13,7 +13,7 @@
 [![Tests: TDD](https://img.shields.io/badge/tests-TDD-brightgreen.svg)](CONTRIBUTING.md)
 [![Quality Gate Status](https://sonarcloud.io/api/project_badges/measure?project=ffroliva_gflow-cli&metric=alert_status)](https://sonarcloud.io/summary/new_code?id=ffroliva_gflow-cli)
 
-> ⚠️ **Read this before you install.** gflow-cli is **unofficial, alpha, and reverse-engineered — not affiliated with Google**. It drives a headed browser on *your own* Google Flow session, so treat it as your own account risk: automation is subject to Google's ToS, and endpoints or UI can change without notice. It requires a Google AI **Ultra or Pro** subscription with Flow access, and every generation bills your account. Read the full [DISCLAIMER](DISCLAIMER.md).
+> ⚠️ **Read this before you install.** gflow-cli is **unofficial, alpha, and reverse-engineered — not affiliated with Google**. It drives a headed browser on *your own* Google Flow session, so treat it as your own account risk: automation is subject to Google's ToS, and endpoints or UI can change without notice. It works with **any Google account** that has Flow access, and every generation bills against your account's Flow credit allowance. Read the full [DISCLAIMER](DISCLAIMER.md).
 >
 > 💳 **What failure costs you.** Credits are only spent on Veo *video* generation — images and composition ops are free, so most breakage costs nothing. When Flow's UI drifts mid-run, the CLI fails fast and loudly with distinct exit codes (e.g. selector drift = exit 23) instead of resubmitting, and batch items are recorded locally *before* submission so a broken run never silently burns credits on a stale state. See [KNOWN_ISSUES](KNOWN_ISSUES.md) for the current risk list.
 >
@@ -21,7 +21,7 @@
 
 ## Why gflow-cli?
 
-You pay for Google AI Ultra or Pro, you have Veo credits, and you run real batch work. gflow-cli gives you:
+You have a Google account with Flow access, you have Veo credits, and you run real batch work. gflow-cli gives you:
 
 - **Batch generation.** Loop prompts straight from the shell: `for p in $(cat prompts.txt); do gflow image t2i "$p"; done`. Image batching plus `gflow video t2v` / `i2v` / `r2v` all ship today.
 - **Consistent subjects.** `gflow character create` mints a Flow Character (face and body reference) so the same person appears from one generation to the next.
@@ -29,7 +29,7 @@ You pay for Google AI Ultra or Pro, you have Veo credits, and you run real batch
 - **Pipelines.** Wire Veo into your content automation, AI-video stack, or batch experiments.
 - **Terminal-native.** After one `gflow auth login`, you stay in the shell. No clicking through dialogs.
 
-Same Veo and Imagen models, same quality, same Ultra/Pro billing, now programmatic.
+Same Veo and Imagen models, same quality, same billing against your own Google account, now programmatic.
 
 ## 60-second quick start
 
@@ -73,7 +73,7 @@ One command in, real Flow output back. Left: `gflow image t2i` generating a phot
 
 ![gflow image t2i runs a single 9:16 prompt, streams structlog output, and writes a PNG to disk](https://raw.githubusercontent.com/ffroliva/gflow-cli/main/docs/assets/example-run.gif)
 
-A single `gflow image t2i "..." --aspect 9:16 --model nano2` call against a logged-in Pro/Ultra profile. The terminal streams the run's `structlog` JSON, then lists the written PNG. Chromium drives the Flow editor silently in the background.
+A single `gflow image t2i "..." --aspect 9:16 --model nano2` call against a logged-in Flow profile. The terminal streams the run's `structlog` JSON, then lists the written PNG. Chromium drives the Flow editor silently in the background.
 
 Reproduce the recording with [`scripts/record_demo.ps1`](scripts/record_demo.ps1) (Windows, OBS, ffmpeg, gifski). More formats, including the side-by-side split-screen: **[docs/DEMOS.md](docs/DEMOS.md)**.
 
@@ -131,7 +131,7 @@ Full milestone history lives in [CHANGELOG.md](CHANGELOG.md). Where the project
 
 ## License & legal
 
-[MIT License](LICENSE) © 2026 Flavio Oliva (`ffroliva`). The MIT license covers `gflow-cli`'s code only. It grants no rights to Flow, Veo model output, or any Google service. Google's own terms (Labs Additional Terms, Ultra/Pro subscription terms) govern your generations. See the [DISCLAIMER](DISCLAIMER.md).
+[MIT License](LICENSE) © 2026 Flavio Oliva (`ffroliva`). The MIT license covers `gflow-cli`'s code only. It grants no rights to Flow, Veo model output, or any Google service. Google's own terms (Labs Additional Terms and any plan-specific subscription terms) govern your generations. See the [DISCLAIMER](DISCLAIMER.md).
 
 ## Acknowledgements
 

{gflow_cli-0.54.0 → gflow_cli-0.56.0}/docs/AUTHENTICATION.md

@@ -10,7 +10,7 @@ This page documents the full lifecycle: capture, storage, reuse, refresh, multi-
 |---|---|
 | Re-implement Google's OAuth dance | Google's web SSO involves anti-automation challenges (CAPTCHA, device verification). A community SDK can't reliably ship that. |
 | Extract the bearer token from cookies and use `httpx` directly | Tokens are short-lived and tied to a refresh flow we don't have. Re-implementing the refresh is brittle. |
-| Use a service-account JSON | Flow doesn't currently support service accounts on the AI Ultra/Pro consumer tier. |
+| Use a service-account JSON | Flow doesn't currently support service accounts on the consumer tier. |
 | Pure stored cookie jar (no Playwright) | Some Flow endpoints set additional `x-` headers based on Chromium fingerprint; matching them by hand is fragile. |
 | **Playwright persistent context (chosen)** | Captures session once, reuses indefinitely, refreshes automatically on idle, auto-attaches every header Flow expects. One-time cost: a ~150 MB Chromium download. |
 
@@ -197,18 +197,20 @@ accidental interference or data corruption.
 
 ### `gflow auth status`
 
-Reports whether a profile exists, where it lives, and whether the cookies file is present.
+Reports where the profile lives, then **proves the session**: it probes the Flow session endpoint with the profile's cookies (no browser, no credits) and exits `0` only when the session is verified — `1` otherwise. Handy for scripts: `gflow auth status && gflow run ...`.
 
 ```bash
 $ gflow auth status
-Profile 'default' is configured.
   profile: /home/you/.local/share/gflow-cli/profile_default
   exists: True
   cookies_present: True
   cookies_path: /home/you/.local/share/gflow-cli/profile_default/Default/Cookies
+  browser_engine: playwright
+Probing Flow session (may take up to ~45s on a slow network)...
+Flow session verified as you@example.com.
 ```
 
-> Note: `cookies_present: True` only confirms the file exists — not that the session is still valid with Google. The first real API call (e.g. `gflow image t2i`) is the actual probe. If Google has invalidated the session, the call will fail with an authentication error and you'll be prompted to re-run `auth login`.
+> Note: `cookies_present: True` only confirms the file exists. The final verdict line is the live probe result: a dead session prints a `gflow auth login` remediation hint and exits 1; a probe that cannot reach the endpoint (offline, 5xx) also exits 1 but suggests checking connectivity instead of re-logging in.
 
 ### `gflow auth list`
 
@@ -367,6 +369,38 @@ Run: gflow auth login --profile default
 
 Re-running `auth login` refreshes the cookies in place — no other state is lost.
 
+## Chromium downgrade guard
+
+Chrome profiles are not backward-compatible: opening a profile with an **older**
+Chromium major version than the one that last wrote it triggers Chromium's
+downgrade cleanup, which can leave the newer store — session cookies included —
+unreadable. That surfaces later as a mystery logout after a gflow/Playwright
+downgrade (or a Chrome-strategy profile silently falling back to an older
+bundled Chromium).
+
+Every bundled-Chromium open of a persisted profile (the generation client, the
+UI-automation transport, the experimental transports, and the headless
+verification probe) first compares the profile's `Last Version` file against
+the active engine's bundled Chromium (`playwright`, or `patchright` when
+selected) and **refuses to open on a major-version downgrade** (exit 11) with
+an error naming both versions:
+
+```text
+ERROR: Profile at ~/.gflow/profile_default was last written by Chromium 150.x,
+but the installed Playwright bundled Chromium is older (149.x). ...
+```
+
+Remedies: upgrade so the bundled Chromium is at least the profile's version
+(`uv tool upgrade gflow-cli`, then `playwright install chromium`), or re-create
+the profile with `gflow auth login` — login is deliberately *not* guarded, it
+re-mints the session and rewrites the profile, making it the recovery path.
+The guard is best-effort: profiles under the `chrome` strategy (real Google
+Chrome opens them) and any unreadable/unknown version skip the check. One
+surface nuance: the headless verification probe is fail-closed by contract, so
+a refusal there is reported as a verification failure rather than exit 11 —
+the version details still land in the structured log
+(`browser_manager.profile_engine_downgrade`).
+
 ## Session verification (cookie-store fast path)
 
 Since v0.17.0, profile verification (run by `gflow auth login --browser chrome` after
@@ -394,7 +428,7 @@ extraction lives in `gflow_cli.auth.cookies`.
 |---|---|
 | Session file leaked to a public repo | `.gitignore` excludes profile dirs at every layer. Recommended belt-and-braces: keep the gflow-cli home outside the repo (default location via `platformdirs` already does this) and run `git status` before any commit. Automatic in-repo detection is on the backlog (not yet scheduled). |
 | Multi-user shared machine | Profiles live under each user's home dir; OS file permissions (`0700` on POSIX, ACL on Windows) prevent cross-user reads by default. |
-| `gflow-cli` itself becomes malicious | The package is open-source under MIT; pin a version (`uv tool install gflow-cli==0.5.0a1`) and review release diffs before upgrading. |
+| `gflow-cli` itself becomes malicious | The package is open-source under MIT; pin a version (`uv tool install gflow-cli==0.55.0`) and review release diffs before upgrading. |
 | Stolen laptop | Anyone with disk access has your session. Use full-disk encryption (FileVault, BitLocker, LUKS). Consider a dedicated `--profile sandbox` for short-lived experiments. |
 | Sharing a profile between machines | Technically works (copy the profile dir), but Google may flag the device-fingerprint mismatch as suspicious. Re-login on the new machine instead. |
 

{gflow_cli-0.54.0/website → gflow_cli-0.56.0}/docs/CONFIGURATION.md

@@ -109,7 +109,7 @@ Deep setup, verification, and security notes live in
 ### `GFLOW_CLI_PROVIDER`
 
 **What:** Backend to use for generations.
-**Values:** `flow` (default — reverse-engineered, requires Ultra/Pro) | `official` (planned v0.5+ — Phase 5 official Veo SDK swap, will require a Gemini API key).
+**Values:** `flow` (default — reverse-engineered, works with any Google account that has Flow access) | `official` (planned v0.5+ — Phase 5 official Veo SDK swap, will require a Gemini API key).
 **Default:** `flow`
 **CLI override:** none yet — set the env var to switch backends once `official` is wired.
 
@@ -205,6 +205,12 @@ gflow tools run creative-director "cat in space" --json   # check "was_expanded"
   no env var set. If a corporate proxy is configured, add your local gateway to `NO_PROXY`
   or its traffic will be routed through the proxy and fail non-obviously.
 
+### `GFLOW_CLI_DAEMON_TOKEN` (alias: `GFLOW_DAEMON_TOKEN`)
+
+**What:** API token required before `gflow serve` will bind to a non-localhost address (`--host` other than `127.0.0.1`). Without it, non-local binds abort with exit 11.
+**Default:** unset.
+**Security:** stored as a Pydantic `SecretStr` (since v0.55.0), so a `repr()`/`str()`/`model_dump_json()` of the settings object masks it by construction — on top of the existing logging-boundary redaction. Treat it like any credential: set it via `.env`/environment, never commit it.
+
 ### `GFLOW_CLI_AUTH_LOGIN_TIMEOUT`
 
 **What:** Maximum time (seconds) that `gflow auth login` waits for the user to complete the Google sign-in flow in the browser.
@@ -254,6 +260,21 @@ GFLOW_CLI_AUTH_LOGIN_TIMEOUT=120 gflow auth login   # abort after 2 minutes
 **Recommended starting point:** `1` until a concurrent caller lands. Each additional Page would cost ~30–60 MiB of memory on Chromium headless; don't exceed `8` without measuring resident-set size. Cookies and storage state are shared at Context level, so every Page would inherit the signed-in profile for free.
 **Shipped in:** v0.4.0a2.
 
+### `GFLOW_CLI_UPDATE_CHECK`
+
+**What:** Once-a-day best-effort PyPI check that prints a one-line stderr notice when a newer gflow-cli exists. The notice is always served from a local cache (`<GFLOW_CLI_HOME>/update_check.json`); a stale cache refreshes on a background daemon thread whose result feeds the *next* invocation — the check never blocks or fails a command, and a failed poll still counts toward the once-a-day cap.
+**Values:** `1` (default) | `0` to disable
+**Skipped automatically:** in CI (`CI` env var set) and for editable/local-source installs (PEP 610 `direct_url.json` detection) — "upgrade" advice is wrong there.
+**Shipped in:** #479.
+
+### `GFLOW_CLI_LEASE_WAIT_SECONDS`
+
+**What:** How long a command waits for another gflow process to release the profile lease before giving up with `ProfileLockedError` (exit 11). The default `0` keeps the historical fail-fast behavior. With a positive value, the waiter polls the kernel lock (0.5 s cadence) and simply takes over when the current holder — a CLI command or a `gflow serve` daemon task, both of which release at their natural end — finishes. Holders are never interrupted or asked to release early.
+**Values:** `0`–`3600` seconds (fractions allowed)
+**Default:** `0` (fail fast)
+**Note:** Same-process contention always fails fast regardless of this setting — the holder is the same process, so waiting would deadlock. See [KNOWN_ISSUES § Same profile can't be used in parallel](../KNOWN_ISSUES.md#same-profile-cant-be-used-in-parallel).
+**Shipped in:** #478.
+
 ### `GFLOW_CLI_DB_PATH`
 
 **What:** Override the path to the local SQLite operations database.
@@ -338,7 +359,7 @@ GFLOW_CLI_HISTORY_PROMPTS=redacted gflow image t2i "confidential brief"
 **Default:** `true`
 **When to disable:** shared machines where even structural failure metadata should not persist, or scripted runs that must write nothing outside the output dir.
 
-**Privacy:** the automatic JSON artifacts (`manifest.json`, `ui.json`, `network.json`, `browser.json`) are built from an explicit allowlist — no prompts, tokens, cookies, headers, request/response bodies, signed URLs, raw page titles, raw error/console text, or unknown hosts/routes ever enter them. The screenshot is inherently sensitive (it can show your account identity, prompts, and media) and therefore lives under the bundle's `sensitive/` subdirectory — **review it before sharing**. Nothing is ever uploaded; retention is bounded (at most 50 complete bundles / 250 MiB, pruned oldest-first at startup). Raw HAR capture stays separate and strictly opt-in via `GFLOW_CLI_HAR_PATH`.
+**Privacy:** the automatic artifacts (`manifest.json`, `ui.json`, `network.json`, `browser.json`, and the pre-filled `report.md` bug-report template) are built from an explicit allowlist — no prompts, tokens, cookies, headers, request/response bodies, signed URLs, raw page titles, raw error/console text, or unknown hosts/routes ever enter them. The screenshot is inherently sensitive (it can show your account identity, prompts, and media) and therefore lives under the bundle's `sensitive/` subdirectory — **review it before sharing**. Nothing is ever uploaded; retention is bounded (at most 50 complete bundles / 250 MiB, pruned oldest-first at startup). Raw HAR capture stays separate and strictly opt-in via `GFLOW_CLI_HAR_PATH`.
 
 ### `GFLOW_CLI_HAR_PATH`
 

{gflow_cli-0.54.0/website → gflow_cli-0.56.0}/docs/DEBUGGING.md

@@ -45,10 +45,11 @@ historical UI-flake noted in
 [`LIVE_VERIFICATION_v0.7.0.md`](LIVE_VERIFICATION_v0.7.0.md#open-follow-ups)
 (transient overlay swallowed the arrow-forward click).
 
-**Two extra debug-screenshot file names worth searching for in the out dir**
-when the mode switch itself fails: `debug_no_mode_trigger.png` (crop-icon
-dropdown trigger not found) and `debug_no_image_tab.png` (Image tab missing
-inside the open dropdown).
+**Two extra artifact file names worth searching for in the out dir** when the
+mode switch itself fails: `diag_mode_switch_miss.json` (crop-icon dropdown
+trigger not found — a structural DOM signature; this probe writes no
+screenshot) and `debug_no_image_tab.png` (Image tab missing inside the open
+dropdown).
 
 Pipe `--verbose` output through `Select-String` (PowerShell) or `grep` to
 filter:
@@ -87,6 +88,9 @@ object) between the UTC timestamp and a collision-resistant random suffix:
 │                             # category, canonical route, status, duration
 ├── browser.json              # last ≤100 console warn/error + ≤50 page errors —
 │                             # class/category + length only, never message text
+├── report.md                 # pre-filled Markdown bug report: version/platform,
+│                             # error class + exit code, artifact pointers — fill
+│                             # in "What I was doing" and attach when filing
 ├── sensitive/screenshot.png  # UI-state failures only — REVIEW BEFORE SHARING
 └── .pending                  # present only while a capture is still staging
 ```
@@ -181,7 +185,7 @@ without manual copying. Caveats verified during the v0.7.0 work:
 ### 3. Firecrawl (`firecrawl-scrape` / `firecrawl-interact`)
 Best for unauthenticated docs (e.g., Material Symbols icon names, third-
 party API references). Not ideal for Flow itself — Flow requires a
-logged-in Pro/Ultra session that Firecrawl's hosted browser doesn't
+logged-in Flow session that Firecrawl's hosted browser doesn't
 share with your local profile.
 
 ### 4. Playwright trace viewer