A lightweight Chrome extension + OpenClaw workflow for turning what you read on the internet into social posts and blog drafts.

This project is built around a simple idea:

Capture while learning → structure automatically → ship faster.

This system has two capture modes:

• Clip mode → saves small insights from reading
• Blog mode → accumulates structured learning for long-form writing

Both modes store content as markdown files, which are later used by an OpenClaw skill to generate drafts.

Chrome Extension
   ├── Clip Mode → clips/*.md
   └── Blog Mode → clips/dossiers/<topic>.md

OpenClaw Skills
   ├── ship_post → social draft
   └── blog_draft → blog draft

Everything is local-first and file-based.

No database required.

Clip mode is designed for small learning moments.

1. Highlight text on any webpage
2. Click Save Clip
3. Extension writes a markdown file to:

Downloads/clips/

Each clip looks like:

---
type: clip
source: web
title: "Floating-Point Numbers - Algorithmica"
url: "https://en.algorithmica.org/hpc/arithmetic/float/"
created_at: 2026-02-15T11:47:17+05:30
tags: ["programming"]
---

## Highlight
> Floating-point numbers store significant digits and scale using an exponent.

## Takeaway
Floating point numbers optimize relative error instead of absolute precision.

An OpenClaw skill reads the newest clip and generates a social post draft. Save the skill in .openclaw/skills/ship_post/SKILL.md

Example command:

/ship_post

Output:

Floating-point numbers store precision using mantissa + exponent.

They optimize relative error instead of exact representation, making numerical computation efficient.

https://en.algorithmica.org/hpc/arithmetic/float/

This creates a low-friction learning → sharing loop.

Blog mode is designed for deep learning on a topic.

Instead of creating many files, it maintains one evolving markdown file per topic.

1. Choose a topic (example: floating-point)
2. Highlight text from articles/videos/docs
3. Extension appends structured content
4. Same markdown file is overwritten with updates
5. File downloads locally

Example location:

Downloads/clips/dossiers/floating-point.md

An OpenClaw skill converts the blog markdown into a draft article.

Example:

/dossier floating-point

Output:

• structured explanation
• narrative flow
• draft blog content
• minimal editing required

The goal is:

Learning notes → coherent blog draft

Most people:

• read a lot
• save bookmarks
• forget insights

This system instead creates:

Reading → Capture → Structure → Ship

It reduces friction between learning and publishing.

Early prototype, Demoed at Build India.

This is not a note-taking tool.

It is a learning-to-publishing pipeline.