You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I recently came across banana-claude and really love the "Creative Director" prompt-engineering pipeline you've built. The way you structure the 5-component prompt formula (Subject, Action, Location, Composition, Style) to coax high-quality images out of Gemini is a fantastic pattern.
I wanted to give you a heads-up about a sibling project I am working on called gflow-cli.
What it is: gflow-cli is an unofficial Python CLI that reverse-engineers Google Flow's private REST API to drive both Imagen (text-to-image) and Veo (text-to-video, image-to-video) generations directly from the terminal.
Why there's cool synergy here:
Local Generation Engine: Since gflow-cli drives generations through a warm browser session via your own Pro/Ultra Google account, it provides an alternative to public AI Studio API limits and costs (especially for video/Veo).
Creative Director Prompting: We are currently exploring adding a similar prompt-expansion capability (like your 5-component formula) to gflow-cli to help terminal users generate richer prompts.
MCP Potential: Exposing gflow-cli as an Model Context Protocol (MCP) server is on our project roadmap. In the future, a plugin like banana-claude could theoretically use gflow-cli as a local provider to generate not just images, but full videos/b-roll right inside Claude Code.
Would love to get your thoughts on this, and see if there are opportunities for collaboration or cross-referencing between the projects!
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
Hi @AgriciDaniel,
I recently came across banana-claude and really love the "Creative Director" prompt-engineering pipeline you've built. The way you structure the 5-component prompt formula (Subject, Action, Location, Composition, Style) to coax high-quality images out of Gemini is a fantastic pattern.
I wanted to give you a heads-up about a sibling project I am working on called gflow-cli.
What it is:
gflow-cliis an unofficial Python CLI that reverse-engineers Google Flow's private REST API to drive both Imagen (text-to-image) and Veo (text-to-video, image-to-video) generations directly from the terminal.Why there's cool synergy here:
gflow-clidrives generations through a warm browser session via your own Pro/Ultra Google account, it provides an alternative to public AI Studio API limits and costs (especially for video/Veo).gflow-clito help terminal users generate richer prompts.gflow-clias an Model Context Protocol (MCP) server is on our project roadmap. In the future, a plugin likebanana-claudecould theoretically usegflow-clias a local provider to generate not just images, but full videos/b-roll right inside Claude Code.Would love to get your thoughts on this, and see if there are opportunities for collaboration or cross-referencing between the projects!
Cheers,
Flavio Oliva / @ffroliva
All reactions