Lee Foot

Gist Summary Generator

Create "At a glance" bullet point summaries from articles using AI.

Launch the app → View the source

#about

Structured "**At a glance**" header with asterisk bullet points.

Three length options: Concise (15-20 words), Standard (20-30), Detailed (30-40).

Configurable bullet count: min 2-5, max 3-7.

Direct factual statements (not "this article discusses").

Firecrawl URL fetching.

Streamlit appRequires an API key

Platform
Browser-based (no installation required)
Input
Anthropic or OpenAI API keyContent: paste, URL (with Firecrawl), or CSV/ExcelBullet length and count settings
Output
"At a glance" summary with key theme bullets. Each point is a distinct factual statement.
Category
Content

Use cases

  • Featured snippet optimisation
  • Newsletter article previews
  • Executive summaries
  • Social media content briefs

#features

  • Three bullet lengths: Concise/Standard/Detailed
  • Configurable bullet count (min 2-5, max 3-7)
  • Direct facts (no meta-descriptions)
  • Firecrawl URL fetching
  • Bulk CSV/Excel processing

#how-to-use

  1. Enter AI API key (optional Firecrawl for URLs)
  2. Select bullet length (Concise/Standard/Detailed)
  3. Set min and max bullet count
  4. Paste content, enter URL, or upload file
  5. Download formatted summary

#faq

Is the whole article sent to the model?
Yes. Unlike some of the sibling tools there is no input truncation, so token cost scales directly with article length, and that is the main cost driver. Output is capped at 1,024 tokens, which is ample for a bullet summary.
Why were some rows skipped in bulk mode?
Any row whose content is under 100 characters is skipped with status 'Content too short to summarize'. Rows where the API call failed record the error in the status column instead; there is no automatic retry, so filter on status and re-run failures.
Do I need a Firecrawl key?
Only for the URL Fetch tab. The paste tab and bulk file processing need just an Anthropic or OpenAI key. Firecrawl fetches main content only with ads and base64 images stripped.
Can I join the bulk results back to my original file?
Only via the title: the results download contains just title, summary and status columns, and none of your other columns (URLs, IDs, content) are carried through. Make sure you map a unique title column before running, or you will have nothing to join on.

/contact

Want me to run this for you? I run this tool as a managed service, or build something custom around your data. You get the insights without touching the code.