Epic: Unified output format — PDF/HTML with inline screenshots

[BraedenBDev]

Problem

Currently, PointDev's output is text-only when copied to clipboard. Screenshots are separate base64 data URLs that can only be copied individually. An AI coding agent (Cursor, Copilot, Claude) needs both structured text AND screenshots in a single paste to understand the full context.

Requirements

• Single "Copy All" action that includes text + screenshots
• Format options: PDF, HTML, or Markdown with embedded images
• Screenshots should be inline next to the voice segment / annotation they belong to
• Output should be pasteable into AI chat interfaces that support images
• Consider multimodal clipboard (ClipboardItem with multiple MIME types)

Design Considerations

• PDF: universal but hard to parse programmatically
• HTML: rich but not all AI tools accept it
• Markdown + base64 images: Claude and similar tools can ingest this
• MCP bridge may be the better delivery mechanism long-term (Shareable captures: export/import structured captures across tools and teams #37)

Related

• Part of NLnet Milestone 2 (prompt compiler)
• Depends on intent linking (Intent linking: merge temporally-adjacent voice segments with annotations #40)

🤖 Generated with Claude Code