Google Flow

Generate and edit images for free in Google Flow (flow.google.com) via the browser-skill `bsk` CLI, optimizing the user's brief into a continuity-aware prompt for the project's Agent panel and Nano Banana. Use for Flow image generation, editing, or prompt optimization, including uploaded references. Video is paid and only on explicit request.

9 stars 1 forksImages, Video & Audio

Install

skills CLI (Claude Code, Cursor, Codex and more)
npx skills add Vincent-the-gamer/skills --skill google-flow
Manual install
git clone --depth 1 https://github.com/Vincent-the-gamer/skills.git /tmp/skills
mkdir -p ~/.claude/skills
cp -r /tmp/skills/skills/google-flow ~/.claude/skills/google-flow

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# Google Flow:免费图片生成

Flow 项目页右侧的 **Agent 面板** 用 Nano Banana 生成和改图,默认免费。本 skill 通过 browser-skill 的 `bsk` CLI 操作这个面板;页面通用操作以 browser-skill 为准,这里只写 Flow 的流程和免费/付费边界。

**默认只做图片。** 视频会消耗积分;只有用户明确要求视频,并且你已告知费用、用户确认后,才进入 [付费视频分支](references/video.md)。

**开始前**:`bsk` CLI 和浏览器扩展必须已装好,扩展显示 Connected。缺 CLI 或扩展就停下,按 [AGENT_INSTALL.md](https://raw.githubusercontent.com/Tencent/BrowserSkill/main/AGENT_INSTALL.md) 处理,把装扩展、登录等需要用户的步骤交给他,连接好之后再继续。

## 每次发送前的零花费检查

1. 在**项目页右侧的 Agent 面板**里,不在图片详情页、视频工具或 `工具` 页。
2. 模型或模式是**图片**(Nano Banana),不是 Veo 或其他视频模型。
3. 提示词只描述**静态画面或改图**,没有镜头运动、时长、音频、视频/做成动画等词。
4. 提交按钮没有**本次消耗积分**的标记(0 credits / 免费);只有账户余额图标不算。

任一项为“否”:停下,先问用户,不要发送。

## 0. 检查连接

```sh
bsk --version
bsk doctor
bsk browsers
```

`bsk browsers` 列出多个实例时,先问用户用哪个账号/浏览器 profile,并在启动会话时用 `--browser <id-or-name>` 绑定。

**完成判据**:`bsk doctor` 没有失败项,且确定了要用的浏览器实例。

## 1. 定 brief(图片)

先确认这次是**生成**还是**改图**,再一次问全:

```text
生成还是改图?
主体 / 场景 / 风格 / 构图 / 光线
画幅(16:9 / 9:16 / 1:1)和张数(页面默认 2x)
参考图:绝对路径 + 每张图的作用(人物 / 场景 / 道具 / 风格)
要改的图:哪一张、改什么、保留什么
画面文字:要出现的原文和位置(没有就写“无”)
视频:本次不做(除非用户主动提出)
```

没有偏好时的默认值:新图 16:9、2 张;改图 1 张;Nano Banana 2.1;无文字;不做视频。

用户给的是图片链接而不是本地文件时,先下载到本地;剪贴板粘贴只认本地文件。确认文件存在且是 PNG/JPEG/WebP:macOS/Linux 用 `ls -l <path> && file <path>`,Windows 用 `Get-Item <path>`,装了 ImageMagick 时三平台都可用 `magick identify <path>` 直接看格式和尺寸。

HEIC 等格式先转换(macOS:`sips -s format png in.heic --out out.png`;Windows/Linux:ImageMagick `magick in.heic out.png`)。
Read the full skill on GitHub

Details

Skill name
google-flow
Path
skills/google-flow/SKILL.md
License
Apache-2.0
Repo updated
15 November 2026

FAQ

What does the Google Flow skill do?

Generate and edit images for free in Google Flow (flow.google.com) via the browser-skill `bsk` CLI, optimizing the user's brief into a continuity-aware prompt for the project's Agent panel and Nano Banana. Use for Flow image generation, editing, or prompt optimization, including uploaded references. Video is paid and only on explicit request.

How do I install the Google Flow skill?

skills CLI (Claude Code, Cursor, Codex and more): npx skills add Vincent-the-gamer/skills --skill google-flow. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Google Flow popular?

It lives in Vincent-the-gamer/skills, which has 9 GitHub stars and 1 forks.

Related skills

Slidev

9

Vincent-the-gamer/skills

Create and present web-based slidedecks for developers using Slidev with Markdown, Vue components, code highlighting, animations, and interactive features. Use when building technical presentations, conference talks, code walkthroughs, teaching materials, or developer decks.

Documents & Office Files

Build MCP App

9

Vincent-the-gamer/skills

This skill should be used when the user wants to build an "MCP app", add "interactive UI" or "widgets" to an MCP server, "render components in chat", build "MCP UI resources", make a tool that shows a "form", "picker", "dashboard" or "confirmation dialog" inline in the conversation, or mentions "apps SDK" in the context of MCP. Use AFTER the build-mcp-server skill has settled the deployment model, or when the user already knows they want UI widgets.

AI, LLMs & Agents

Build MCP Server

9

Vincent-the-gamer/skills

This skill should be used when the user asks to "build an MCP server", "create an MCP", "make an MCP integration", "wrap an API for Claude", "expose tools to Claude", "make an MCP app", or discusses building something with the Model Context Protocol. It is the entry point for MCP server development — it interrogates the user about their use case, determines the right deployment model (remote HTTP, MCPB, local stdio), picks a tool-design pattern, and hands off to specialized skills.

AI, LLMs & Agents

Radmin Remote

9

Vincent-the-gamer/skills

Use Radmin Viewer 3 on Windows to connect to a remote computer, and drive its authentication dialog programmatically. Use when the user asks to 远程/remote into a host by IP with Radmin, needs full-control or view-only control, needs a non-default mode (file transfer, telnet, text chat, voice chat, send message, shutdown), or asks which command-line switches Radmin supports.

Images, Video & Audio

Fast Dirpy

9

Vincent-the-gamer/skills

fast-dirpy is a Node.js CLI that extracts direct video links and downloads videos from supported sites. Use when the user says to use fast-dirpy to download a video or get a direct link.

Images, Video & Audio

Waifu2x

9

Vincent-the-gamer/skills

Upscale and denoise anime-style images on waifu2x.net through the browser-skill `bsk` CLI — apply the user's choices, upload a local file or image URL, hand the Cloudflare check to the user, convert, then download and verify the result. Use for anime image upscaling (1.6x/2x), denoising, or converting a picture to PNG/WebP. First run asks for each option and offers to save them in a waifu2x.md.

Images, Video & Audio

Browse the most popular Claude skills, all images, video & audio skills, or write your own with the skills generator.