# AI use cases for Flow MCP (GTAI-1/flow-mcp) Repo: [GTAI-1/flow-mcp](https://github.com/GTAI-1/flow-mcp) | Use cases: 4 | Page: https://dedupedai.com/use-cases/gtai-1-flow-mcp/ flow-mcp is a free, locally run program that drives Google Flow in a dedicated Chrome window so Claude, Codex or a local control panel can make AI images, video and narration using Flow credits instead of the pay-per-use video API. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Make a short AI film from one character: https://dedupedai.com/use-cases/gtai-1-flow-mcp/make-a-short-ai-film-from-a-character/ - Batch-render scene stills at zero credit cost: https://dedupedai.com/use-cases/gtai-1-flow-mcp/batch-render-scene-stills-for-free/ - Archive a Flow project's media locally at no cost: https://dedupedai.com/use-cases/gtai-1-flow-mcp/archive-flow-project-media-for-free/ - Add free narration and join it under your clips: https://dedupedai.com/use-cases/gtai-1-flow-mcp/narrate-and-join-a-voiceover-film/ ## How to make a short AI film from one character with Flow MCP Repo: [GTAI-1/flow-mcp](https://github.com/GTAI-1/flow-mcp) | Difficulty: Intermediate | Time: 120 minutes | Basis: From the README Create a reusable character, draw key stills, then shoot and join short clips into one film. **Who it is for:** A solo creator or developer who wants a short animated scene with one recurring character and has a Google account with Flow access. **The problem:** Making a consistent character across several AI video shots is slow when each clip is generated by hand in Flow's interface. **The outcome:** A finished MP4 in ~/flow-mcp-out/ containing a character-led film built from several Flow clips, joined in order. ### Steps 1. Run `git clone https://github.com/GTAI-1/flow-mcp.git` and then `cd flow-mcp`. 2. Run `npm install` and then `npm run build`. 3. Run `npm run login`, sign into Google by hand in the Chrome window that opens, open or create a Flow project, and leave it open. 4. Run `claude mcp add --scope user flow -- node /absolute/path/to/flow-mcp/dist/index.js` and restart Claude Code. 5. Ask the assistant to check the Flow session, create a character, draw free stills, then shoot the clips with a credit cap before any paid step. ### Prompt to paste into your agent ```text Using the flow-mcp repo at , check the Flow session, create a character named , draw three free stills with that character, and then shoot three 10-second 720p clips with max_credits set to per clip. Expected output is a joined MP4 under ~/flow-mcp-out/. Verify that flow_status reports signed in and that the credit cap was not exceeded before shooting. ``` ### Caveat Shots spend real Flow credits and the Chrome window must stay open and awake, so the run can stop if the display sleeps. Risk, credentials: The tool uses a Chrome profile signed into your Google account at ~/.flow-mcp/chrome-profile, so anyone with access to that folder can use your session. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/gtai-1-flow-mcp/make-a-short-ai-film-from-a-character/ ## How to batch-render scene stills at zero credit cost with Flow MCP Repo: [GTAI-1/flow-mcp](https://github.com/GTAI-1/flow-mcp) | Difficulty: Intermediate | Time: 60 minutes | Basis: From the README Send up to 50 scene prompts to Flow's Agent mode and get matched image files back. **Who it is for:** A storyboard artist or game designer with a list of scene descriptions who needs many reference images quickly without spending credits. **The problem:** Rendering a long list of images one at a time in Flow is slow, and matching results back to their prompts by hand is error-prone. **The outcome:** Up to 50 images rendered in parallel, each saved under the scene number it was given, at 0 credits. ### Steps 1. Run `git clone https://github.com/GTAI-1/flow-mcp.git`, then `cd flow-mcp`, then `npm install` and `npm run build`. 2. Run `npm run login` and sign into Flow in the Chrome window that opens, then open a Flow project. 3. Run `codex mcp add flow -- node /absolute/path/to/flow-mcp/dist/index.js` to connect Codex. 4. Give the assistant a list of up to 50 scene prompts and ask it to use the agent image batch tool. 5. Open ~/flow-mcp-out/ and check that each image file matches its scene number. ### Prompt to paste into your agent ```text Using the flow-mcp repo at , send these scene prompts from to the agent image batch tool in one call. Expected output is one image per scene in ~/flow-mcp-out/ with file numbers that match the scene order. Verify that the returned results list every scene and that the credit count shown is 0. ``` ### Caveat Batches are rendered inside Flow's own Agent mode, so the tool depends on that Flow feature staying available in your project. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/gtai-1-flow-mcp/batch-render-scene-stills-for-free/ ## How to archive a Flow project's media locally at no cost with Flow MCP Repo: [GTAI-1/flow-mcp](https://github.com/GTAI-1/flow-mcp) | Difficulty: Intermediate | Time: 45 minutes | Basis: From the README Download existing images, clips and stitched films from Flow, including 1080p upscales, without regenerating. **Who it is for:** A creator who has built up work in a Flow project and wants a local copy before the project changes or runs out of space. **The problem:** Work made in Flow stays in the web project, and downloading each clip by hand or regenerating it costs time and credits. **The outcome:** Local copies of the project's media in ~/flow-mcp-out/, with each scene saved as one finished film and upscales included. ### Steps 1. Run `git clone https://github.com/GTAI-1/flow-mcp.git`, then `cd flow-mcp`, then `npm install` and `npm run build`. 2. Run `npm run login` and open the Flow project you want to archive in the Chrome window. 3. Run `claude mcp add --scope user flow -- node /absolute/path/to/flow-mcp/dist/index.js` and restart Claude Code. 4. Ask the assistant to list the project's assets, then download them with the free download tool. 5. Check ~/flow-mcp-out/ for the downloaded files. ### Prompt to paste into your agent ```text Using the flow-mcp repo at , list the assets in the open Flow project and download all videos and images with the free download tool, including the 1080p upscale where available. Expected output is local media files in ~/flow-mcp-out/ with one stitched film per scene. Verify that the credits remaining shown by flow_status did not drop during the download. ``` ### Caveat Full-quality downloads depend on your Google plan, since the README says a paid plan unlocks 1080p downloads. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/gtai-1-flow-mcp/archive-flow-project-media-for-free/ ## How to add free narration and join it under your clips with Flow MCP Repo: [GTAI-1/flow-mcp](https://github.com/GTAI-1/flow-mcp) | Difficulty: Intermediate | Time: 45 minutes | Basis: From the README Record narration in a Google or Mac voice, then join it under your clips with music. **Who it is for:** A creator with finished Flow clips who needs a spoken narration and background music in one local MP4. **The problem:** Recording narration and mixing it with several clips usually means separate tools and paid voice services. **The outcome:** One MP4 in ~/flow-mcp-out/ with the clips joined in order, a narration track on top, and the clips' own sound turned down. ### Steps 1. Install FFmpeg with `brew install ffmpeg` on a Mac, or `winget install Gyan.FFmpeg` on Windows. 2. Run `git clone https://github.com/GTAI-1/flow-mcp.git`, then `cd flow-mcp`, then `npm install` and `npm run build`. 3. Run `npm run studio` and open http://127.0.0.1:8787 in your browser. 4. On the Voice tab, record a narration in a Google voice or your computer's own voice. 5. On the Cut tab, pick your clips in order, add the narration and music, and turn the clips' sound down before exporting. ### Prompt to paste into your agent ```text Using the flow-mcp repo at , record the narration text with the free engine, then join the clips in in order under that narration with background music. Expected output is one MP4 in ~/flow-mcp-out/. Verify that the file plays and the narration is audible over the lowered clip sound. ``` ### Caveat Google narration needs a free Google AI Studio key, and the Mac voice option needs no key, so pick the engine that matches your setup. Risk, credentials: Google voices need a Google AI Studio key, which the panel stores on your computer, so keep it out of chats and shared folders. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/gtai-1-flow-mcp/narrate-and-join-a-voiceover-film/