home/marketing/describe-figure

Describe Figure

GPTClaudeGemini··1,358 copies·updated 2026-07-14
describe-figure.prompt
Analyze this image extracted from a document page. First, determine if it is an informative figure worth indexing.

IGNORE and respond with exactly "NO_FIGURE" if the image is any of:
- A full page of scanned text with no visual figures
- A logo, brand mark, watermark, or company emblem
- A stamp, seal, notary mark, or official certification mark
- A signature or handwritten initials
- A decorative element, border, separator, or icon
- A header, footer, or page number region
- Any image that is purely ornamental and carries no informative visual content

DESCRIBE only if the image is an informative figure such as:
- A chart, graph, or plot (bar, line, pie, scatter, etc.)
- A table with structured data
- A diagram, flowchart, or schematic
- A map or architectural/engineering drawing
- A photograph or illustration that conveys information relevant to the document content

If it is an informative figure, describe what it visually shows in 2-4 concise sentences. Focus on the visual content: what objects, charts, or diagrams are depicted, their spatial arrangement, and any key data or measurements visible. Do not repeat surrounding document text or captions.
{context}

fill the variables

This prompt has 1 variable. Pro fills them into a ready-to-paste prompt for you — no manual find-and-replace.

{context}
Unlock with Pro →

when to use it

Community prompt sourced from the open-source GitHub repo hericlesferraz/DocVault (MIT). A "Describe Figure" style prompt — adapt the placeholders and specifics to your task. Imported as-is and not independently retested here, so check the output before relying on it.

tags

marketingcommunitygeneral

source

hericlesferraz/DocVault · MIT