moar turns docs into AI-ready Markdown, CSV

Hot take: this is less about “compression” and more about making document inputs legible to LLMs without forcing users into a workflow tax.

• –The strongest angle is the AI-native framing: optimize for model comprehension, not human readability.
• –Local-only processing is a real differentiator for confidential docs and enterprise-adjacent use cases.
• –The “up to 95% token savings” claim is attention-grabbing, but the practical value is whether the cleaned output preserves structure consistently across messy real-world files.
• –Supporting multiple formats and large files makes it feel more like an input pipeline than a one-off converter.
• –The Chrome-extension form factor lowers friction, which matters more here than flashy feature breadth.