Reference boards
Prompts, references, and results in one place
Start with the limitation, because it decides whether this is for you. A board only knows what you put in it. Scopeful does not read your generation history from Midjourney or Kling and reconstruct your past work. You save things, and then they stay useful. If you want something that organizes work you have already scattered across six tools, this is not that.
What is a reference board?
A reference board is a single place that holds the prompts, the reference images, and the finished results for one piece of work. The link between the three is the point. Stored on its own, a prompt is a string that once worked. Six months later you have no record of what it was aimed at, which reference set the look, or whether the output was any good. A board keeps all three attached, so the question “is this worth reusing” has an answer you can see rather than one you have to remember.
How is a reference board different from a prompt library?
A prompt library is organized around text you wrote. A reference board is organized around work you did. The difference is invisible on day one and obvious after a few months. A library hands you a string to paste. A board hands you the string, the reference it came from, and the output it produced, which is enough to tell whether it still holds up or needs rewriting for a model that shipped since.
This matters more than it used to. Image and video models change fast enough that a prompt written for last year’s version of a tool often reads wrong to the current one. Without the reference and the result stored alongside it, you cannot tell the difference between a prompt that never worked and a prompt the model outgrew.
What does the Copilot do with a board?
It reads the references on the board and writes a prompt phrased for the specific tool you name. The same board produces a different prompt for Midjourney than for Kling, because the two take different parameters and respond to different phrasing. Add a different reference and the prompt changes with it. You are not writing prompt syntax from memory for each tool; you are pointing at the work and naming the destination.
Which tools does it write prompts for?
Scopeful tracks 40 AI creative tools, including Midjourney, Kling, Krea, and Runway. Each one has a tool page with pricing verified by hand and normalized to USD per generation, so a credit bundle and a monthly subscription can be compared as one number. Every page lists the source URL the pricing came from and the date it was last checked. You can verify the figure instead of trusting it.
Can an AI coding agent use this?
Yes, through the Model Context Protocol. The @scopeful/mcpserver exposes the pricing engine to any MCP client, including Claude Code, Cursor, and Windsurf. A key created with write access also reaches the user’s own boards, saved prompts, and reference images, so an agent can save a prompt to a board the same way the web app does. The free tier allows 100 MCP calls per month. The Starter plan at 20 dollars per month raises that to 3,000.
Start a board
Boards live in the app. Pricing, comparisons, and the tool catalog stay on this site and stay free.