Go from an idea to a finished image or clip without leaving the chat
You describe what you want, your agent makes it, and you refine it in the same thread. No second tab, no copy and paste.
Cancel anytime before day 7 · no lock-in
Why leaving the chat slows you down
Here is the usual routine. You have an idea in your chat, you open a separate image site, you sign in, you paste the prompt, you wait, you download the result, then you come back and describe your change somewhere else. Every switch costs a little attention.
Figgshield removes the switch. It is a small MCP server at https://api.figgshield.ai/mcp that your AI agent calls whenever you ask for a picture or a video. The generator sits inside the assistant you already talk to.
So the whole job happens in one place. You ask, you look, you adjust, and the result shows up in the same conversation.
A first run, from prompt to file
Say you want a simple product shot. Here is what one round looks like, start to finish.
That is the entire loop. Each round is one more message, not one more app to open.
- Describe it in plain words: a flat-lay photo of a ceramic coffee cup on a linen cloth, with soft morning light.
- Your agent calls Figgshield, which runs Nano Banana 2 from Google to make the image.
- A private link comes back in the chat and you take a look.
- Ask for the change in the same message: warmer light, and move the cup to the left.
- It makes a new version, and you keep the one you like.
Refine without starting over
Because everything lives in one thread, you refine instead of restarting. You do not re-explain the brief or re-upload anything, since the agent already has it.
Every still it makes has an id, so you can point back at an earlier one. Say go back to the second version and darken the background, and it picks up the image you meant.
A separate tab cannot do that. When the generator and the conversation share a place, your earlier tries are still there to build on.
When a still is not enough, ask for motion
You can turn a picture you already made into a moving clip by passing its id. The coffee cup you liked becomes a five-second shot with steam rising from it.
Video runs on Kling 3, with a faster Kling 3 Turbo tier when you want the result sooner. A five-second clip is 35 credits.
Since the chat already holds the image id, you do not upload anything again. You say animate the last image, gentle steam, camera holds still, and it uses what you have.
What it costs, in plain numbers
One credit balance covers both stills and clips. An image is 8 credits at 1K or 2K, and 16 credits at 4K. A five-second clip is 35 credits.
Pricing is at cost. We pass the model price straight through with no markup. The only money Figgshield keeps is on credits people buy and never spend.
You can see current rates on the pricing page and top up on the credits page.
Try it on a real idea
It works with Claude on desktop and in Claude Code, with Cursor and Windsurf, and with ChatGPT in Developer Mode on a paid plan. Any MCP agent can call it.
There is a 7-day Pro trial that includes 150 credits for $5. There is no free tier. Cancel before it renews and you pay nothing more.
Pick one small idea you keep meaning to make, and make it in the next chat you open.