get-tldr_skill

Overview

This skill provides the exact summary returned by the get-tldr.com summarize API and formats it for readable output without changing its content. It extracts the full value of the response's "summary" property and returns it verbatim, preserving any markdown. The skill ensures the summary is not wrapped in a code block so markdown renders correctly.

How this skill works

When triggered with a URL, the skill runs the bundled get_tldr.py script which posts {"input": "<URL>"} to the get-tldr.com API using an API key from config or environment. The script prints the JSON response to stdout; the skill reads the response, takes the entire "summary" field, and outputs it exactly as received. If the summary is already markdown, it returns that markdown unchanged (except for removing a surrounding code block wrapper if present).

When to use it

• You want the API-provided summary exactly as the service produced it, with no additional summarization or interpretation.
• You need the summary text preserved verbatim for compliance, archival, or replication of results.
• You want markdown output from the API to render correctly in your client (headers, lists, links).
• You are automating link summarization where deterministic, machine-produced output is required.

Best practices

• Provide a single, well-formed URL as the argument so the API can fetch the target content reliably.
• Install and configure the API key at ~/.config/get-tldr/config.json or set GET_TLDR_API_KEY to avoid failures.
• Do not attempt to re-summarize, paraphrase, or redact the returned summary — deliver it verbatim.
• If you display the summary in a UI, ensure you allow markdown rendering and avoid wrapping the text in code blocks.
• Log API responses only to configured logfile locations to respect privacy and audit needs.

Example use cases

• A chat agent receives a URL from a user and returns the official TL;DR produced by get-tldr.com for quick consumption.
• Archiving summaries of news articles or blog posts while preserving the original machine-generated wording.
• Comparing outputs from different summarization services by storing their exact summaries for side-by-side review.
• Feeding the API-produced markdown summary into a renderer that can display headings, lists, and links as intended.