summary-distiller

Solid

USE WHEN the user pastes long text (article, paper, transcript, doc) and wants a summary. Picks the right summary shape (TL;DR / key points / structured / actionable) for the source type and length. Preserves numbers, names, dates verbatim.

AI & Automation 50 stars 6 forks Updated today MIT

Install

View on GitHub

Quality Score: 80/100

Stars 20%
57
Recency 20%
100
Frontmatter 20%
70
Documentation 15%
100
Issue Health 10%
50
License 10%
100
Description 5%
100

Skill Content

# summary-distiller — useful summaries, not lossy ones ## Pick the shape based on source | Source | Shape | |--------|-------| | News article | TL;DR + 3-5 key facts + "why it matters" | | Academic paper | Background / Method / Results / Limitations | | Meeting transcript | Decisions / Action items (with owner+due) / Open questions | | Tutorial / how-to | Numbered steps, preserving every command verbatim | | Long thread / comments | Stance map: "X argued A, Y countered B" | | Strategy doc | Goals / Constraints / Decisions / Open issues | Don't impose one shape on all sources. If unsure, ask: "What are you using this for?" — that tells you the shape. ## Rules ### Preserve verbatim These must be quoted exactly, never paraphrased: - **Numbers** — "$1.2M", "37%", "3,400 users" - **Dates** — preserve the original date format - **Names** of people, products, places, papers - **Quotes** explicitly attributed to a speaker - **Commands / code** — copy character-for-character ### Compress aggressively elsewhere - Cut adjectives and qualifiers ("very", "really", "essentially") - Merge redundant statements - Drop background the audience already knows ### Length targets - TL;DR: 1 sentence, ≤25 words - Key points: 3-5 bullets, ≤20 words each - Structured summary: ≤25% of source word count - Quotes section: only direct quotes that matter, max 5 ## Common failure modes - Summarizing the **structure** instead of the **content** ("the article discusses three points") — the user...

Details

Author
hesorchen
Repository
hesorchen/muselab
Created
2 months ago
Last Updated
today
Language
Python
License
MIT

Integrates with

Similar Skills

Semantically similar based on skill content — not just same category