plan-from-lucid

plan-from-lucid

Fetch a Lucidchart diagram and use its content as context for /speckit.plan.

Arguments

$ARGUMENTS contains: <lucid-url> [additional-input]

• lucid-url (required): URL to a Lucidchart document page, e.g. https://lucid.app/lucidchart/9c028565-3df6-4c34-9db4-449bb5c7c20d/edit?page=4WDSdv3BMKuG
• additional-input (optional): Extra context or instructions to include alongside the diagram

Steps

1. Parse Arguments

Extract the Lucid URL (first whitespace-delimited token from $ARGUMENTS) and any additional input (everything after it).

2. Parse the Lucid URL

From the URL, extract:

• documentId: the GUID path segment between /lucidchart/ and /edit, e.g. 9c028565-3df6-4c34-9db4-449bb5c7c20d
• pageId: the value of the page query parameter, e.g. 4WDSdv3BMKuG

If pageId is absent from the URL, omit it when calling the fetch tool (the server will default to the first page).

3. Fetch the Diagram

Using the Lucidchart MCP server, call its document fetch tool with documentId and pageId. Request the content in the most structured format available — prefer SVG or XML over raster images, as structured vector formats are more accurately interpreted by the AI. If the server returns a PNG image, use it as-is for visual analysis.

4. Compose Context

Format the retrieved diagram content into a context block:

Lucidchart Document: <documentId>
Page: <pageId>

Diagram:
<diagram content — SVG, XML, or image as returned by the MCP server>

Append any additional input from $ARGUMENTS after the above block.

5. Invoke /speckit.plan

Pass the composed context block as the feature description to /speckit.plan.