{"slug": "show-hn-1pager-fixing-the-ai-verbosity-problem", "title": "Show HN: 1pager: Fixing the AI verbosity problem", "summary": "A developer released 1pager, an AI skill that condenses long documents, directories, or chats into a strict one-page summary with bullet-first formatting and minimal verbosity, addressing the common problem of verbose AI output. The tool exports to Markdown and .docx, and aims to reduce the 'validation tax' on readers by forcing brevity and removing AI-writing tells.", "body_md": "A skill that condenses any long document, directory, or chat into a strict\n**one-page** summary — bullet-first, least verbosity possible — humanizes it, and\nexports a `.md`\n\nplus a `.docx`\n\n(opens in Word and Google Docs).\n\nLLMs are verbose by default. Left unedited, their output is long, padded, and full of tells (rule-of-three, \"delve\", em-dash sprees, generic upbeat closers).\n\nThe cost lands on the reader. As Camille Fournier argues in\n[Guidelines for respectful use of AI](https://skamille.medium.com/guidelines-for-respectful-use-of-ai-affcc85d7072),\nshipping unreviewed, bloated AI output creates a **validation tax**: colleagues\nhave to read and verify text the author never bothered to trim. Long PRs, wall-\nof-text messages, and bloated docs are, in her words, \"frankly, just rude.\"\nBrevity is a form of respect — shorter is better because it protects the\nreader's time.\n\nThis skill exists to do the trimming the author should have done: cut the output down to only what the reader needs to understand or decide, on a single page.\n\n**Strips to essentials**— keeps the point, the decision/ask, the few supporting facts, real risks, and concrete next steps. Drops everything else.**Drafts a strict 1-pager**— bottom line first, then bullets. One page, hard ceiling (~350 words).** Humanizes**— removes AI-writing tells. Invokes the installed`humanizer`\n\nskill if present; otherwise falls back to the bundled`references/humanizer.md`\n\n.**Exports two files**—`<name>-1pager.md`\n\nand`<name>-1pager.docx`\n\n.\n\n```\n1pager/\n├── SKILL.md                     # the skill: condense → humanize → export\n├── README.md                    # this file\n├── scripts/\n│   └── md_to_docx.py            # pure-stdlib Markdown → .docx (no dependencies)\n└── references/\n    └── humanizer.md             # bundled humanizer rules (fallback)\n```\n\nIn Claude Code, the skill triggers when you ask to condense, shorten, or make a 1-pager out of something:\n\n\"Condense this design doc into a one-pager.\" \"This thread is way too long — give me a tl;dr I can share.\" \"Boil the docs/ folder down to a single page.\"\n\nPoint it at a **directory** and it combines the relevant files into one page by\ndefault (one page per file if you ask for that instead).\n\nYou can also run the converter directly:\n\n```\npython3 scripts/md_to_docx.py my-1pager.md my-1pager.docx\n```\n\n`md_to_docx.py`\n\nsupports the small Markdown subset a 1-pager needs: `#`\n\n–`####`\n\nheadings, `-`\n\n/`*`\n\nbullets (two-space indent = sub-bullet), `**bold**`\n\n, and\n``code``\n\n(rendered monospace). It has no external dependencies — a `.docx`\n\nis\njust a zip of XML, built with the Python standard library.", "url": "https://wpnews.pro/news/show-hn-1pager-fixing-the-ai-verbosity-problem", "canonical_source": "https://github.com/cfitzgerald-pd/1pager/tree/main", "published_at": "2026-06-18 18:19:18+00:00", "updated_at": "2026-06-18 18:31:33.558136+00:00", "lang": "en", "topics": ["ai-tools", "large-language-models", "developer-tools"], "entities": ["Camille Fournier"], "alternates": {"html": "https://wpnews.pro/news/show-hn-1pager-fixing-the-ai-verbosity-problem", "markdown": "https://wpnews.pro/news/show-hn-1pager-fixing-the-ai-verbosity-problem.md", "text": "https://wpnews.pro/news/show-hn-1pager-fixing-the-ai-verbosity-problem.txt", "jsonld": "https://wpnews.pro/news/show-hn-1pager-fixing-the-ai-verbosity-problem.jsonld"}}