gflow-cli 0.74.0__tar.gz → 0.75.0__tar.gz

{gflow_cli-0.74.0 → gflow_cli-0.75.0}/.agents/plugins/marketplace.json

@@ -8,7 +8,7 @@
       "name": "gflow",
       "source": {
         "source": "local",
-        "path": "./"
+        "path": "./plugins/gflow"
       },
       "policy": {
         "installation": "AVAILABLE",

gflow_cli-0.75.0/.claude-plugin/marketplace.json

@@ -0,0 +1,17 @@
+{
+  "name": "gflow-cli",
+  "owner": {
+    "name": "Flavio Oliva",
+    "url": "https://github.com/ffroliva"
+  },
+  "description": "Drive Google Flow (Veo video, Imagen images) from Claude Code.",
+  "plugins": [
+    {
+      "name": "gflow",
+      "source": "./plugins/gflow",
+      "category": "Developer Tools",
+      "tags": ["video", "image", "google-flow", "veo", "imagen", "mcp"],
+      "defaultEnabled": false
+    }
+  ]
+}

{gflow_cli-0.74.0 → gflow_cli-0.75.0}/.codex-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "gflow",
-  "version": "0.74.0",
+  "version": "0.75.0",
   "description": "Reusable development and operations workflows for gflow-cli",
   "author": {
     "name": "Flavio Oliva",
@@ -14,7 +14,7 @@
     "codex",
     "google-flow"
   ],
-  "skills": "./skills/",
+  "skills": "./plugins/gflow/skills/",
   "interface": {
     "displayName": "gflow-cli",
     "shortDescription": "Develop, validate, review, and operate gflow-cli.",

{gflow_cli-0.74.0 → gflow_cli-0.75.0}/.github/workflows/ci.yml

@@ -199,6 +199,9 @@ jobs:
       - name: Council memory (every cited slug resolves, every file is cited)
         run: uv run python scripts/ci/check_council_memory.py
 
+      - name: Plugin skills in sync (curated payload, no maintainer-only skill shipped)
+        run: uv run python scripts/ci/generate_plugin_skills.py --check
+
       - name: Lint
         run: uv run ruff check src tests
 

{gflow_cli-0.74.0 → gflow_cli-0.75.0}/AGENTS.md

@@ -44,7 +44,7 @@ by construction, so Codex / Cursor / Aider / `agy` read exactly what Claude Code
 
 ## Project at a glance
 
-- 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` — and, for accounts Google has moved to `flow.google.com`, that frontend's `batchexecute` wire (text-to-video, image-to-video from a local `--initial-frame`, reference-to-video from local `--ref` files, and text-to-image/local-file image-to-image, today; `GFLOW_CLI_FLOW_HOST`).
+- Python CLI and MCP server 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` — and, for accounts Google has moved to `flow.google.com`, that frontend's `batchexecute` wire (text-to-video, image-to-video from a local `--initial-frame`, reference-to-video from local `--ref` files, and text-to-image/local-file image-to-image, today; `GFLOW_CLI_FLOW_HOST`).
 - 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`, `diagnostics.py`, `json_output.py`, `media.py`, `profile_lease.py`, `redaction.py`, `storage.py`, `winsec.py`, `cli_project.py`, `cli_character.py`, `cli_credits.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`, `cli_doctor.py`, `cli_update.py`, `composition.py`, `config.py`, `errors.py`, `file_integrity.py`, `flow_selectors/`, `update_check.py`, `exceptions.py`, `image_batch.py`, `movie_manifest.py`, `observability.py`, `paths.py`, `profile_store.py`.
 - Command surface: `gflow auth`, `gflow credits` (user/list — read-only Veo balance), `gflow image` (t2i/i2i/batch/upload/upscale), `gflow video` (t2v/i2v/r2v/chain/extend — `extend` continues an existing clip past Flow's 8s ceiling, server-seeded from the source so the join is continuous; 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 doctor` (read-only pre-flight diagnostic, exit 33 = findings present), `gflow update` (self-update through the installer that put it here — uv tool / pipx / pip; `--check` only reports; source installs refused, exit 11; deliberately no MCP twin), `gflow project`, `gflow models`, `gflow run`, `gflow mcp` (run/setup — stdio MCP server), and `gflow serve` (Streamable HTTP at `/mcp`; `--transport sse` is deprecated).

{gflow_cli-0.74.0 → gflow_cli-0.75.0}/CHANGELOG.md

@@ -7,6 +7,93 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [0.75.0] — 2026-09-15
+
+### Added
+
+- **Installable as a Claude Code plugin.** `/plugin marketplace add ffroliva/gflow-cli` then
+  `/plugin install gflow@gflow-cli` installs the `gflow-cli` and `video-production` skills and
+  registers the MCP server in one step. The plugin ships **disabled**: Claude Code starts a
+  plugin's MCP servers automatically on enable with no prompt of its own, and this one drives your
+  own Google account where video generation bills your credits, so enabling it is a deliberate act.
+  For a hard guarantee, register the server yourself with `gflow mcp run --no-spend`.
+- **`docs/DISTRIBUTION.md`** — an operational catalog of every channel people install or discover
+  gflow-cli through: audience, how to submit, requirements, status and a last-verified date per
+  row. Every row was checked live. It also records the channels that listed us without being asked
+  (MCP Market, skills.sh with 21 installs, two auto-generated catalogs) and the ones we are not
+  eligible for, with the reason.
+- **Listed on [cursor.directory](https://cursor.directory/plugins/gflow-cli)**, and submitted to
+  seven more channels: Glama, mcpservers.org, and PRs or issues on `punkpeye/awesome-mcp-servers`,
+  `ComposioHQ/awesome-claude-skills`, `hesreallyhim/awesome-claude-code`,
+  `travisvn/awesome-claude-skills` and `Arnon-hs/open-source`. `DISTRIBUTION.md` carries the
+  status and reference for each, and — equally deliberately — why four channels are closed to a
+  local-stdio server that spends the user's own credits, including a section on why there is no
+  hosted "connector" and what one would actually cost.
+- **`server.json`** — metadata for the official MCP Registry, with `tests/test_server_json.py`
+  pinning version lockstep, the schema's 100-character description cap, the name pattern, the
+  `mcp-name:` ownership token in the README, and that the command it advertises really exists.
+- **A `gflow-cli` console script.** `uvx gflow-cli mcp run` previously failed with uv's own
+  *"Use `uvx --from gflow-cli <EXECUTABLE-NAME>` instead"*, because the console scripts were named
+  `gflow` and `flow`. The MCP Registry builds exactly that `uvx <identifier>` command from the
+  PyPI identifier and has no field for a differing executable name, so a listing would have been
+  broken on arrival.
+
+### Changed
+
+- **PyPI metadata.** The summary described only image-to-video and never mentioned MCP; it now
+  says what the package is. Added `Documentation`, `Repository` and `Changelog` sidebar links,
+  ten trove classifiers (all checked against the official list) and MCP-related keywords.
+- **Dropped "unofficial" as a label** from the PyPI summary, `README.md`, `index.html`, the docs
+  site, `llms.txt`, `ROADMAP.md`, `DISCLAIMER.md`, `AGENTS.md`, `CLAUDE.md`,
+  `skills/gflow-cli/SKILL.md` and `src/gflow_cli/__init__.py`. The substance is unchanged and
+  still prominent: the README warning block reads "alpha and reverse-engineered — not affiliated
+  with Google", and `DISCLAIMER.md` still opens "not affiliated with, endorsed by, sponsored by,
+  or otherwise connected to Google LLC" — only the word "unofficial" left its first sentence.
+  It was leading with a negative in the one line PyPI shows in search results.
+
+### Fixed
+
+- **The Codex and ChatGPT-desktop plugin manifests shipped every skill in `skills/`**, including
+  maintainer-only ones (`release`, `check`, `pr-council-review`, `sonar`, `doc-review`). All three
+  channels now point at the same curated two-skill payload under `plugins/gflow/`, generated from
+  `skills/` by `scripts/ci/generate_plugin_skills.py` with a `--check` drift gate in CI.
+
+- **Remediation hints naming an extra rendered without the extra.** Rich reads `[chain]` in an
+  interpolated value as a style tag and silently drops it, so `pip install 'gflow-cli[chain]'`
+  printed as `pip install 'gflow-cli'` — advice that reinstalls what you already have. The same
+  applied to `gflow-cli[patchright]`. Only brackets whose first character is `[a-z#/@]` are
+  affected, which is exactly the shape of a package extra. Every site that renders error text
+  through Rich now escapes it, and `tests/test_rich_markup_safety.py` fails the build if a new one
+  appears. `--json` was never affected.
+
+- **`gflow video chain` was unusable on a clean `gflow-cli[chain]` install (#813).** The extra
+  installed `av` but not Pillow, while `media.py` imports `PIL` at module level — so
+  `uvx --from 'gflow-cli[chain]==0.74.0' gflow video chain one.jsonl --dry-run` died on
+  `No module named 'PIL'`, surfaced as a generic exit `1` *"Unexpected error… file a bug"*,
+  immediately after the user installed the documented extra. `pillow>=12.3.0` now ships in the
+  `chain` extra alongside `av`.
+- **A missing `av` used to fail only AFTER a paid clip.** `import av` sat inside
+  `media._decode_frame`, which runs between links — so the missing extra surfaced once link 0
+  had already been generated and billed. It is now a module-level import beside `PIL`, and the
+  import block at the top of `video chain` is guarded: either missing package is a typed
+  `FrameExtractionError` (exit `20`) raised *before* the manifest is read, before `--dry-run`
+  prints a plan, and before the cost prompt. Nothing is submitted and no browser is launched.
+  `FrameExtractionError`'s remediation now names the extra **and both** packages (`av`,
+  `pillow`); it previously said only "PyAV", which left a missing Pillow undiagnosable.
+- **Error text containing `[...]` was silently truncated on the console.** `detail` and
+  `remediation_hint` were interpolated into Rich markup unescaped, so Rich read `[chain]` as an
+  unknown style tag and dropped it — turning every `install gflow-cli[chain]` hint into
+  `install gflow-cli`, advice that reinstalls what the user already has. Both error render
+  paths now escape. `--json` was never affected.
+### Security
+
+- **`gflow serve` now requires the configured daemon token on every HTTP request.**
+  `GFLOW_CLI_DAEMON_TOKEN` / `GFLOW_DAEMON_TOKEN` previously gated startup only. It is now verified
+  on every request, on both the Streamable HTTP and the deprecated SSE transport, with a
+  constant-time comparison; a missing, malformed or wrong token gets `401` plus a
+  `WWW-Authenticate: Bearer` header and never reaches a tool. Set a token and upgrade if you run
+  `gflow serve` on anything other than the default loopback bind.
+
 ## [0.74.0] — 2026-09-14
 
 ### Added
@@ -5014,7 +5101,8 @@ shell-script template that branches on these codes.
 
 First skeleton. Not functional end-to-end yet.
 
-[Unreleased]: https://github.com/ffroliva/gflow-cli/compare/v0.74.0...HEAD
+[Unreleased]: https://github.com/ffroliva/gflow-cli/compare/v0.75.0...HEAD
+[0.75.0]: https://github.com/ffroliva/gflow-cli/compare/v0.74.0...v0.75.0
 [0.74.0]: https://github.com/ffroliva/gflow-cli/compare/v0.73.2...v0.74.0
 [0.73.2]: https://github.com/ffroliva/gflow-cli/compare/v0.73.1...v0.73.2
 [0.73.1]: https://github.com/ffroliva/gflow-cli/compare/v0.73.0...v0.73.1

{gflow_cli-0.74.0 → gflow_cli-0.75.0}/CLAUDE.md

@@ -4,7 +4,7 @@
 
 ## What this project is
 
-`gflow-cli` is an unofficial Python CLI that drives [Google Flow](https://labs.google/fx/tools/flow) (Veo image-to-video, Imagen text-to-image) from the terminal by reverse-engineering Flow's private REST API. See [README.md](README.md) for the user-facing overview.
+`gflow-cli` is a Python CLI that drives [Google Flow](https://labs.google/fx/tools/flow) (Veo image-to-video, Imagen text-to-image) from the terminal by reverse-engineering Flow's private REST API. See [README.md](README.md) for the user-facing overview.
 
 ## System rules — loaded, not requested
 

{gflow_cli-0.74.0 → gflow_cli-0.75.0}/DISCLAIMER.md

@@ -2,7 +2,7 @@
 
 ## Not affiliated with Google
 
-`gflow-cli` is an **independent, unofficial** project. It is **not affiliated with, endorsed by, sponsored by, or otherwise connected to Google LLC, Alphabet Inc., DeepMind, or any of their subsidiaries.** All trademarks (Google, Flow, Veo, Gemini, Imagen, AI Ultra, AI Pro) are the property of their respective owners. The use of these names here is purely descriptive — to identify the service this tool integrates with — and does not imply endorsement.
+`gflow-cli` is an **independent** project. It is **not affiliated with, endorsed by, sponsored by, or otherwise connected to Google LLC, Alphabet Inc., DeepMind, or any of their subsidiaries.** All trademarks (Google, Flow, Veo, Gemini, Imagen, AI Ultra, AI Pro) are the property of their respective owners. The use of these names here is purely descriptive — to identify the service this tool integrates with — and does not imply endorsement.
 
 ## Reverse-engineered surface
 

{gflow_cli-0.74.0 → gflow_cli-0.75.0}/KNOWN_ISSUES.md

@@ -1203,6 +1203,42 @@ your prompts.
 
 ---
 
+### `gflow-cli[chain]` ≤ 0.74.0 does not install Pillow — `video chain` fails with exit 1
+
+- **Status:** **Fixed in v0.75.0** · affects `gflow-cli` ≤ 0.74.0 · **Severity:** High (the command was unusable) · **Affects:** `gflow video chain` · **Tracked:** [#813](https://github.com/ffroliva/gflow-cli/issues/813)
+
+Through 0.74.0 the `chain` extra declared `av` alone, but `gflow_cli/media.py`
+imports `PIL` at module level. So the documented install produced a CLI that
+could not run the command it was installed for:
+
+```console
+$ uvx --isolated --from 'gflow-cli[chain]==0.74.0' gflow video chain one.jsonl --dry-run
+Unexpected error ... file a bug          # exit 1; the real cause is No module named 'PIL'
+```
+
+A missing **`av`** was worse than it looked: `import av` was deferred into the
+decode helper, which only runs *between* links — so it surfaced after link 0 had
+already been generated and **paid for**.
+
+**Workaround on ≤ 0.74.0** — install Pillow alongside the extra:
+
+```bash
+pip install 'gflow-cli[chain]' pillow
+# or:  uv tool install 'gflow-cli[chain]' --with pillow
+```
+
+**Fixed in v0.75.0:** `pillow` ships in the `chain` extra, `av` moved to
+a module-level import so both fail at the same point, and `video chain` now
+raises `FrameExtractionError` (**exit 20**) naming the extra and both packages —
+before the manifest is read, before `--dry-run` prints a plan, and before the
+cost prompt. Verified against a real wheel installed without extras: 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.
+The advice is also legible now — Rich used to eat `[chain]` out of it and print
+`pip install 'gflow-cli'`, i.e. reinstall what you already have.
+
+---
+
 ### `gflow video chain` outputs N clips, not one file — auto-concat is deferred
 
 - **Status:** Open (by design) · **Severity:** Low · **Affects:** `gflow video chain` (v0.12.0)

{gflow_cli-0.74.0 → gflow_cli-0.75.0}/PKG-INFO

@@ -1,8 +1,11 @@
 Metadata-Version: 2.5
 Name: gflow-cli
-Version: 0.74.0
-Summary: Unofficial CLI for Google Flow — drive Veo image-to-video generations from the terminal.
+Version: 0.75.0
+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/
+Project-URL: Repository, https://github.com/ffroliva/gflow-cli
+Project-URL: Changelog, https://github.com/ffroliva/gflow-cli/blob/main/CHANGELOG.md
 Project-URL: Issues, https://github.com/ffroliva/gflow-cli/issues
 Project-URL: Funding, https://github.com/sponsors/ffroliva
 Author-email: Flavio Oliva <ffroliva@gmail.com>
@@ -28,15 +31,25 @@ License: MIT License
         OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
         SOFTWARE.
 License-File: LICENSE
-Keywords: ai,cli,flow,google,veo,video
+Keywords: ai,ai-video,cli,flow,generative-ai,google,google-flow,image-to-video,imagen,mcp,mcp-server,model-context-protocol,playwright,text-to-video,veo,video,video-generation
 Classifier: Development Status :: 3 - Alpha
 Classifier: Environment :: Console
+Classifier: Framework :: AsyncIO
+Classifier: Framework :: Pydantic :: 2
 Classifier: Intended Audience :: Developers
+Classifier: Intended Audience :: End Users/Desktop
 Classifier: License :: OSI Approved :: MIT License
+Classifier: Natural Language :: English
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: Python :: 3 :: Only
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.13
+Classifier: Topic :: Multimedia :: Graphics
 Classifier: Topic :: Multimedia :: Video
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Classifier: Topic :: Utilities
+Classifier: Typing :: Typed
 Requires-Python: >=3.11
 Requires-Dist: browser-cookie3>=0.20.1
 Requires-Dist: click>=8.1.0
@@ -53,6 +66,7 @@ Requires-Dist: tenacity>=8.2
 Requires-Dist: uvicorn[standard]>=0.34.0
 Provides-Extra: chain
 Requires-Dist: av>=12; extra == 'chain'
+Requires-Dist: pillow>=12.3.0; extra == 'chain'
 Provides-Extra: dev
 Requires-Dist: av>=12; extra == 'dev'
 Requires-Dist: numpy>=1.24; extra == 'dev'
@@ -70,9 +84,11 @@ Requires-Dist: s3fs>=2024.2.0; extra == 's3'
 Requires-Dist: universal-pathlib>=0.2.5; extra == 's3'
 Description-Content-Type: text/markdown
 
+<!-- mcp-name: io.github.ffroliva/gflow-cli -->
+
 # gflow-cli
 
-> Unofficial Python CLI for Google Flow. Drive [Veo](https://labs.google/fx/tools/flow) (image-to-video, text-to-video) and Imagen (text-to-image) from your terminal: scripted, batched, pipeline-ready.
+> Python CLI and MCP server for Google Flow. Drive [Veo](https://labs.google/fx/tools/flow) (image-to-video, text-to-video) and Imagen (text-to-image) from your terminal: scripted, batched, pipeline-ready.
 
 [![PyPI version](https://img.shields.io/pypi/v/gflow-cli.svg)](https://pypi.org/project/gflow-cli/)
 [![CI](https://github.com/ffroliva/gflow-cli/actions/workflows/ci.yml/badge.svg)](https://github.com/ffroliva/gflow-cli/actions/workflows/ci.yml)
@@ -92,7 +108,7 @@ Description-Content-Type: text/markdown
 
 <!-- sponsors-gold:end -->
 
-> ⚠️ **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).
+> ⚠️ **Read this before you install.** gflow-cli is **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).
 >
 > 🛡️ **"Will this get my account flagged?"** The honest, specific answer — what the tool does to stay unremarkable (headed real Chrome, randomised interaction timing, paced submissions), what it deliberately does **not** do (no proxies, no fingerprint spoofing, no pretending it isn't automation), what you can tune, and what we cannot promise — is in [docs/ACCOUNT_SAFETY.md](docs/ACCOUNT_SAFETY.md).
 >
@@ -188,6 +204,25 @@ gflow-cli ships four agent entry points. Pick the one your tool reads first.
 | [**llms.txt**](llms.txt) | LLM-readable summary (llmstxt.org format) | Paste into ChatGPT, Claude, or Gemini to onboard the model |
 | [`skills/gflow-cli/SKILL.md`](skills/gflow-cli/SKILL.md) | Claude Code Skill | Symlink into `~/.claude/skills/` |
 
+### Install the plugin (Claude Code)
+
+One step, and you get the `gflow-cli` and `video-production` skills plus the MCP server:
+
+```
+/plugin marketplace add ffroliva/gflow-cli
+/plugin install gflow@gflow-cli
+```
+
+The plugin ships **disabled**. Claude Code starts a plugin's MCP servers automatically once a
+plugin is enabled, with no prompt of its own — and this server drives your own Google account,
+where Veo video generation bills your credits. So installing it starts nothing, and enabling it
+is a deliberate act. Images and composition are free; only video spends. For a hard guarantee,
+register the server yourself with `gflow mcp run --no-spend`, which never registers the
+credit-spending tools at all. See [docs/MCP.md](docs/MCP.md) for the details, including which
+revision `/plugin marketplace add` gives you.
+
+Codex users: `codex plugin marketplace add .` then `codex plugin add gflow@gflow-cli`.
+
 Onboard any agent in one line. Paste this into your agent of choice:
 
 > *"Read [AGENTS.md](https://github.com/ffroliva/gflow-cli/blob/main/AGENTS.md) and [docs/INDEX.md](https://github.com/ffroliva/gflow-cli/blob/main/docs/INDEX.md), then help me with my Flow batch."*

{gflow_cli-0.74.0 → gflow_cli-0.75.0}/README.md

@@ -1,6 +1,8 @@
+<!-- mcp-name: io.github.ffroliva/gflow-cli -->
+
 # gflow-cli
 
-> Unofficial Python CLI for Google Flow. Drive [Veo](https://labs.google/fx/tools/flow) (image-to-video, text-to-video) and Imagen (text-to-image) from your terminal: scripted, batched, pipeline-ready.
+> Python CLI and MCP server for Google Flow. Drive [Veo](https://labs.google/fx/tools/flow) (image-to-video, text-to-video) and Imagen (text-to-image) from your terminal: scripted, batched, pipeline-ready.
 
 [![PyPI version](https://img.shields.io/pypi/v/gflow-cli.svg)](https://pypi.org/project/gflow-cli/)
 [![CI](https://github.com/ffroliva/gflow-cli/actions/workflows/ci.yml/badge.svg)](https://github.com/ffroliva/gflow-cli/actions/workflows/ci.yml)
@@ -20,7 +22,7 @@
 
 <!-- sponsors-gold:end -->
 
-> ⚠️ **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).
+> ⚠️ **Read this before you install.** gflow-cli is **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).
 >
 > 🛡️ **"Will this get my account flagged?"** The honest, specific answer — what the tool does to stay unremarkable (headed real Chrome, randomised interaction timing, paced submissions), what it deliberately does **not** do (no proxies, no fingerprint spoofing, no pretending it isn't automation), what you can tune, and what we cannot promise — is in [docs/ACCOUNT_SAFETY.md](docs/ACCOUNT_SAFETY.md).
 >
@@ -116,6 +118,25 @@ gflow-cli ships four agent entry points. Pick the one your tool reads first.
 | [**llms.txt**](llms.txt) | LLM-readable summary (llmstxt.org format) | Paste into ChatGPT, Claude, or Gemini to onboard the model |
 | [`skills/gflow-cli/SKILL.md`](skills/gflow-cli/SKILL.md) | Claude Code Skill | Symlink into `~/.claude/skills/` |
 
+### Install the plugin (Claude Code)
+
+One step, and you get the `gflow-cli` and `video-production` skills plus the MCP server:
+
+```
+/plugin marketplace add ffroliva/gflow-cli
+/plugin install gflow@gflow-cli
+```
+
+The plugin ships **disabled**. Claude Code starts a plugin's MCP servers automatically once a
+plugin is enabled, with no prompt of its own — and this server drives your own Google account,
+where Veo video generation bills your credits. So installing it starts nothing, and enabling it
+is a deliberate act. Images and composition are free; only video spends. For a hard guarantee,
+register the server yourself with `gflow mcp run --no-spend`, which never registers the
+credit-spending tools at all. See [docs/MCP.md](docs/MCP.md) for the details, including which
+revision `/plugin marketplace add` gives you.
+
+Codex users: `codex plugin marketplace add .` then `codex plugin add gflow@gflow-cli`.
+
 Onboard any agent in one line. Paste this into your agent of choice:
 
 > *"Read [AGENTS.md](https://github.com/ffroliva/gflow-cli/blob/main/AGENTS.md) and [docs/INDEX.md](https://github.com/ffroliva/gflow-cli/blob/main/docs/INDEX.md), then help me with my Flow batch."*

{gflow_cli-0.74.0 → gflow_cli-0.75.0}/RELEASE.md

@@ -90,6 +90,8 @@ python -m pip install --pre gflow-cli
 6. Update the version in:
    - `pyproject.toml`
    - `.codex-plugin/plugin.json`
+   - `plugins/gflow/.claude-plugin/plugin.json` (the Claude Code plugin a user installs;
+     `tests/test_plugin_manifests.py::test_plugin_version_tracks_pyproject` fails if it lags)
    - `src/gflow_cli/__init__.py`
    - `uv.lock` (run `uv lock` — the editable package version is pinned there)
    - any tests that assert the package version
@@ -97,7 +99,7 @@ python -m pip install --pre gflow-cli
    release section.
 8. Commit the release prep:
    ```bash
-   git add pyproject.toml .codex-plugin/plugin.json src/gflow_cli/__init__.py uv.lock CHANGELOG.md tests docs
+   git add pyproject.toml .codex-plugin/plugin.json plugins/gflow/.claude-plugin/plugin.json src/gflow_cli/__init__.py uv.lock CHANGELOG.md tests docs
    git commit -m "chore(release): vX.Y.Z"
    ```
 9. Tag and push. **Must be a signed annotated tag** (`-s`) — `.github/workflows/release.yml` rejects unsigned or lightweight tags. Requires a GPG or SSH signing key registered with your GitHub account.

{gflow_cli-0.74.0 → gflow_cli-0.75.0}/ROADMAP.md

@@ -1,6 +1,6 @@
 # Roadmap
 
-> Themes, not deadlines. `gflow-cli` is an unofficial CLI tracking a private Google Flow API — a single upstream UI change can rearrange a sprint, so we publish themed milestones with deliverables and let velocity emerge.
+> Themes, not deadlines. `gflow-cli` is an independent CLI tracking a private Google Flow API — a single upstream UI change can rearrange a sprint, so we publish themed milestones with deliverables and let velocity emerge.
 
 ## v0.9.0 — Maturity & Visibility (current)
 
@@ -43,4 +43,4 @@ The point at which `FlowApiClient` carries a SemVer commitment and the documenta
 
 ---
 
-*Themes, not deadlines. An unofficial CLI tracking a private API can have its sprint broken by a single UI change upstream — fixed dates would be dishonest.*
+*Themes, not deadlines. An independent CLI tracking a private API can have its sprint broken by a single UI change upstream — fixed dates would be dishonest.*

{gflow_cli-0.74.0 → gflow_cli-0.75.0}/docs/CONFIGURATION.md

@@ -207,8 +207,8 @@ gflow tools run creative-director "cat in space" --json   # check "was_expanded"
 
 ### `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.
+**What:** API token for `gflow serve`. When set, **every** HTTP request to the daemon must carry `Authorization: Bearer <token>` — missing, wrong, or non-`Bearer` credentials are answered `401` with a `WWW-Authenticate: Bearer` header, on both `--transport http` (`/mcp`) and the deprecated `--transport sse` (`/sse` and `/messages/`). It is also 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 — a loopback bind then serves unauthenticated requests (the local single-user posture), and startup logs `mcp.server.auth_disabled` saying so.
 **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_MCP_NO_SPEND`
@@ -476,8 +476,8 @@ each link in turn, so total wallclock is the sum of all link waits).
 | `--profile NAME` | default profile | Per-subcommand profile override. |
 | `--json` | off | Emit a machine-readable JSON result. |
 
-The last-frame extractor needs the **`chain` optional extra** (PyAV — no system
-ffmpeg required):
+The last-frame extractor needs the **`chain` optional extra** (PyAV for decoding —
+no system ffmpeg required — plus Pillow to write the seed JPEG):
 
 ```bash
 pip install 'gflow-cli[chain]'