Overview

MCP Prompts are parameterised message templates that MCP servers expose. They let you capture common workflows — code review patterns, data analysis templates, document generators — and invoke them with a few inputs.

Prompts view with mcp server selected listing 3 prompts

Select a connection to load its prompts. Each prompt shows its name, description, and argument count.

Executing a Prompt

  1. Click a prompt to open its detail panel
  2. Fill in any required arguments
one_sentence_summary prompt with text argument filled in as ms aspire

Each prompt argument has a label, description, and validation indicator. Fill in the values and click Execute.

  1. Click Execute — the rendered MCP messages appear as JSON in the result panel
Prompt result showing rendered output with Send to LLM and Copy buttons

After execution, the result is displayed inline. The Send to LLM button appears in the result toolbar — click it to pipe the prompt messages directly to your configured LLM.

  1. Click Send to LLM — a model picker popover appears
Model picker popover with GPT-4o selected and Cancel/Send buttons

Choose a model from the dropdown (your configured LLM providers are listed here — GPT-4o is the default) and click Send to stream the response.

  1. The LLM Response tab activates with the streamed result
LLM Response tab showing the GPT-4o generated answer to the prompt

The LLM Response tab shows the streamed reply from GPT-4o. Results are rendered as Markdown with syntax highlighting.

Prompt Arguments

Arguments are generated dynamically from the prompt’s schema. Each field shows:

  • A label and description (from the server schema)
  • Validation (required vs optional)
  • Inline error messages

Document Generation

The summary_benefits_and_references prompt demonstrates the document generation capability — it produces a full Markdown document with a summary, key benefits, and references for any topic.

summary_benefits_and_references prompt selected with topic argument field Document output showing formatted Markdown with summary, bullets, and references

The document prompt generates structured Markdown with a summary paragraph, benefit bullets, and reference list — ready to copy or pipe to an LLM.

Favourites

Star any prompt to mark it as a favourite. Toggle Show favourites first in the toolbar to pin starred prompts to the top of the list. This preference is persisted across sessions.

Markdown Rendering

Prompt output is rendered as Markdown, including:

  • Code blocks with syntax highlighting
  • Tables
  • Inline formatting

Send to Chat

After executing a prompt, the Send to Chat button opens the Chat view with the prompt result pre-loaded as the conversation context — letting you continue the interaction with any configured LLM.