HackerNews Extract

**Judgement:** A focused and useful utility skill for converting Hacker News threads (article + comments + metadata) into a single, clean Markdown file. It is well-specified for agent use and fits naturally into LLM and RAG workflows that need context-rich web discussion data. **Strengths:** - Consolidates article, metadata, and threaded comments into one Markdown artifact, ideal for downstream processing. - Clear agent workflow: run via `uv`, save to `/tmp/hn-<id>.md`, then upload and message in a single tool call. - Handles multiple input formats (ID, URL, JSON) and auto-manages dependencies via `uv`. - Includes HTTP retries, which should improve robustness against transient network issues. **Risks / Limitations:** - Depends on `uv` being available in the environment; in constrained or custom runtimes this may reduce compatibility. - Scraping may fail for sites that require authentication or aggressively block bots, leading to partial or missing article content (though HN comments should still work when HN is reachable). - No built-in safety filtering: harmful, offensive, or sensitive content from articles or comments will be passed through verbatim in the Markdown. - Functionality is narrow (HN-focused); not a general-purpose web extractor. **Recommended Scenarios:** - Agents tasked with: “extract this Hacker News post/article and prepare it for analysis/summarization.” - Workflows that want to keep chat clean: the agent should upload the generated Markdown file and only summarize or quote from it when specifically requested. - RAG or research pipelines that benefit from having both the source article and the HN discussion in a single, structured text file. - Power users or analysts who regularly read HN and want a standardized, offline-compatible Markdown view of threads for later study or note-taking.