MCP server for AI assistants
Claude, Cursor, ChatGPT and other MCP clients connect with the server URL and act as the signed-in user.
AI and automation
AI and automation for digital signage, with limits you set.
Most signage work is routine. Find the offline players, refresh the events list, switch on the alert. Wallboard lets AI assistants, other systems and plain-language prompts do that work. Each one acts through the same permissions as the person or key behind it, and admins can switch AI access off for the whole organization.
How it works
How AI fits into Wallboard.
An admin sets the limits
Turn AI access on or off for the organization, and set what each user’s assistant may read, change or delete.
Someone describes the task
Ask an assistant to list offline devices, or type a filter such as outdated Samsung devices into the filter panel.
The request becomes normal actions
The assistant calls the same API a person’s session would. The filter generator fills in the standard filter form.
People review and apply
Generated filters wait in a preview until someone applies them. Blocked MCP actions come back as a clear denial.
What ships today.
AI filter generator
Describe a filter for devices, content or files. The AI turns it into rules you can edit before applying.
AI web scraper
Give the web scraper a public URL and a prompt. It returns the page’s content as structured JSON on a schedule.
AI Assistant for help
A chat in the console answers questions from the Wallboard knowledge base, using your page and role as context.
Webhook actions
A scoped API key lets another system trigger an action, such as emergency content, a datasource change or a refresh.
AI endpoints for your apps
Developers can call Wallboard’s AI API for text, images and spoken audio, billed against your AI budget.
AI on the Proof of Play page.
On the Proof of Play page the AI button sets the page’s own controls instead of building a filter. Type plays per day for the last seven days, or break this down by content, and a preview lists the time range, measured value, breakdown and chart type it will set. Nothing changes until you apply it. Apply and explain also describes what the new view shows. The page saves nothing, and every control stays editable afterwards.
Related pages.
Connect to LLMs
Set up Claude, ChatGPT or Cursor with Wallboard’s MCP server, and decide what each user’s assistant can do.
Skills guide
Write prompts and saved instructions that give the same safe result every time you run them.
AI scraping
Turn event listings, menus or job boards on public websites into data your screens can show.
APIs, MCP and custom applications
The REST API and webhook endpoints that sit behind every feature on this page.
Docs and guides.
Start with one routine task.
Name a task your team repeats every week. In the demo we’ll show which part an assistant, a webhook or a prompt can take over, and our support team can help you set the permissions.