Actions you tune by hand
Your agent opens an action in the action studio. Drag the sliders until the preview looks right, then apply.

Works in Claude, ChatGPT, Codex and Cursor. 50+ models, including Seedance, Kling, Veo and Nano Banana.






























You'll paste this URL into Claude in the next step.
In the Claude desktop app or claude.ai, open Customize → Connectors, then + Add → Add custom connector. Name it FLORA, paste the URL and click Add. It also works in Claude Code sessions in the desktop app.
Click Connect on FLORA and sign in, then open Claude with the prompt ready to generate a campaign deck.
Requires steps 1-2 completed first.
FLORA is listed in the ChatGPT apps directory. Open the FLORA plugin in ChatGPT to install it.
Follow the prompts in ChatGPT to sign in to FLORA and connect.
Once connected, ask ChatGPT to create with FLORA. Try this prompt with a product image.
Adds the FLORA marketplace, then the plugin with the MCP server and every FLORA skill.
Sign in, then ask Claude Code to generate a campaign deck.
Drop this into ~/.cursor/mcp.json, or the project's .cursor/mcp.json.
Cursor signs in through the browser the first time it calls a FLORA tool.
Paste into Cursor’s chat.
Two commands: add the server, then sign in through the browser.
The prompt points Codex at the skill, then at the job.
Paste into Codex’s chat.
Everything your agent does happens where you can see it, and you can take over at any step.
Your agent opens an action in the action studio. Drag the sliders until the preview looks right, then apply.

Your agent loads a skill by name, like PDP Creator or Storyboarding, and runs it start to finish.
Every step lands as a node, wired to the last. Open it and keep going by hand.
In a FLORA Labs session, Claude built a pancake mix brand on a FLORA canvas, then took notes and revised it.
A prompt box returns one result and forgets it. With FLORA, your agent works the way your team does.
Image, video, audio, text and 3D, including Seedance, Kling, Veo and Nano Banana. Run up to 20 generations at once.
Save a workflow once in FLORA and your agent runs it again on new inputs.
Your agent places and connects nodes on a real FLORA canvas, so every step stays editable.
Prebuilt actions reframe an image for any placement, grade it or trim a video. None of them spend credits.
Your agent reads comments on the canvas, replies to them and revises the work.
Every generation is quoted in US dollars before it runs, then billed to your workspace.
Give your agent a photo or a brief, and the skill runs every step through to finished work.

One image becomes a matched series of halftone posters.
Pull anything apart like melted cheese.

A cover from a brief, then the vinyl, the gig poster and the merch.

A pencil sketch becomes a real product, then a campaign.
A product photo becomes a slow turntable video.

One key visual becomes merch, a shop window and a sleeve.
Add FLORA as a custom connector in Claude (Settings, Connectors, Add custom connector, paste https://agents.flora.ai/mcp), sign in with your FLORA account, then ask Claude to generate a video. FLORA quotes the cost in US dollars before anything runs. Setup for every client is at docs.flora.ai/mcp/install.
The same 50+ models as the FLORA canvas, including Seedance, Kling, Veo, Nano Banana, Flux and Seedream for image and video, plus audio, text and 3D models. Your agent lists them with flora_list_models and picks one for each generation.
Connecting is free. A Free FLORA account can generate on the usage balance every new workspace gets, and API and MCP access is included from the Starter plan up. Each generation is billed to your workspace in US dollars and quoted before it runs.
No. FLORA MCP signs you in with OAuth using your existing FLORA account. Headless clients can use a FLORA API key instead, documented at docs.flora.ai/mcp/authentication.