| name | preview |
| description | Render, summarize, and present markdown documents and structured content in multiple output modes |
| layer | hub |
| category | workflow |
| triggers | ["/preview","preview this","show me the doc","render this markdown","summarize this document","present this"] |
| inputs | [{"source":"File path, URL, or inline content to preview"},{"mode":"full | summary | outline | diff | presentation (optional, defaults to full)"},{"audience":"Who the preview is for -- technical, executive, user (optional)"}] |
| outputs | [{"renderedContent":"The document presented in the requested mode"},{"metadata":"Document stats (word count, sections, reading time)"},{"suggestions":"Improvement recommendations for the document (optional)"}] |
| linksTo | ["research","scout"] |
| linkedFrom | ["cook","team","ship","plan","plan-archive"] |
| preferredNextSkills | ["code-review","plan"] |
| fallbackSkills | ["research"] |
| riskLevel | low |
| memoryReadPolicy | selective |
| memoryWritePolicy | selective |
| sideEffects | ["Reads files","Does NOT modify any files"] |
Preview Skill
Purpose
Render, summarize, and present documents and structured content in formats optimized for different audiences and purposes. This skill transforms raw documents into consumable presentations without altering the source material.
Preview is the read-only presentation layer. It helps people understand documents quickly, verify content before publishing, and adapt material for different audiences.
Workflow
Mode 1: Full Preview
Renders the complete document with enhanced formatting and metadata.
- Read the source -- Load the file, URL content, or inline text.
- Analyze the structure:
- Count sections, subsections, paragraphs
- Identify document type (technical doc, plan, report, tutorial, etc.)
- Calculate reading time (avg 200 words/minute)
- List all headings as a navigation aid
- Present the full content with:
- Table of contents (for documents with 3+ sections)
- Section dividers for visual clarity
- Code blocks properly identified by language
- Tables properly aligned
- Links validated (exist or broken)
- Append metadata footer:
---
Words: [count] | Sections: [count] | Reading time: [N] min
Last modified: [date if available]
Mode 2: Summary
Produces a concise summary of the document, preserving key information.
- Read the full document.
- Identify the document's purpose -- What is the main point?
- Extract key points -- One bullet per major section.
- Produce a summary with:
- One-line TL;DR: The document's essence in a single sentence
- Key points: 3-7 bullet points covering the main content
- Action items: Any tasks, decisions, or follow-ups mentioned
- Notable details: Important specifics (numbers, dates, names) that should not be lost
- Tailor to audience:
- Technical: Preserve technical details, code references, architecture decisions
- Executive: Focus on outcomes, timelines, risks, decisions needed
- User: Focus on what changed, what they need to do, what they need to know
