home / skills / ratacat / claude-skills / medium-paywall-bypass

medium-paywall-bypass skill

This skill helps you access paywalled Medium articles by automatically using Freedium and fallbacks to extract readable content.

npx playbooks add skill ratacat/claude-skills --skill medium-paywall-bypass

Review the files below or copy the command above to add this skill to your agents.

Files (1)

SKILL.md

2.5 KB

---
name: medium-paywall-bypass
description: Use when user shares a Medium article URL behind a paywall and wants to read the full content. Also use for articles on Medium-hosted publications like towardsdatascience.com, betterprogramming.pub, levelup.gitconnected.com, etc.
---

# Medium Paywall Bypass

## Overview
Fetch paywalled Medium articles using free mirror services. Try services in order until one works.

## Service Priority

| Service | URL Pattern | WebFetch | curl | Notes |
|---------|-------------|----------|------|-------|
| **Freedium** | `https://freedium.cfd/{encoded_url}` | Yes | Yes | Best option, returns content directly |
| **Archive.today** | `https://archive.today/latest/{raw_url}` | No | Maybe | Often requires captcha |
| **RemovePaywalls** | `https://removepaywalls.com/{raw_url}` | No | No | Redirect page only, needs browser |
| **ReadMedium** | `https://readmedium.com/en/{encoded_url}` | No | No | Returns 403 programmatically |

- `{encoded_url}` = URL-encoded (slashes become %2F, @ becomes %40, etc.)
- `{raw_url}` = Original URL as-is

**For Claude Code: Use Freedium via WebFetch.** Other services require browser interaction.

## Workflow

```
1. User provides Medium URL
2. Try Freedium first via WebFetch
3. If blocked/empty, try next service
4. Extract and present article content
```

## Example Usage

Given: `https://medium.com/@user/some-article-abc123`

**WebFetch (recommended):**
```
URL: https://freedium.cfd/https%3A%2F%2Fmedium.com%2F%40user%2Fsome-article-abc123
Prompt: Extract the full article content
```

**curl fallback:**
```bash
curl -sL "https://freedium.cfd/https%3A%2F%2Fmedium.com%2F%40user%2Fsome-article-abc123"
```

## Medium-Hosted Domains

These domains use Medium's paywall system:
- `medium.com`, `*.medium.com`
- `towardsdatascience.com`
- `betterprogramming.pub`
- `levelup.gitconnected.com`
- `javascript.plainenglish.io`
- `uxdesign.cc`
- `hackernoon.com`
- `codeburst.io`
- `itnext.io`
- `proandroiddev.com`
- `infosecwriteups.com`

## Common Issues

| Problem | Solution |
|---------|----------|
| Freedium down | Try alternative mirror: `freedium-mirror.cfd` |
| Article not found | Article may be too new to be cached |
| Garbled HTML | Use WebFetch with prompt: "Extract the article text and format as markdown" |
| 403/blocked | Try curl with `dangerouslyDisableSandbox: true` |

## Quick Reference

```python
# URL encoding in Python
from urllib.parse import quote
encoded = quote(url, safe='')

# For WebFetch tool
freedium_url = f"https://freedium.cfd/{quote(medium_url, safe='')}"
```

Overview

This skill fetches full Medium articles that are behind Medium's metered paywall by trying free mirror services in priority order. It prefers a direct mirror that returns article HTML and falls back to alternatives when the first option is blocked or empty. The skill extracts and returns the article text in a readable format.

How this skill works

Given a Medium or Medium-hosted publication URL, the skill constructs mirror service URLs (URL-encoding where required) and requests them in sequence until a usable response is obtained. The recommended path uses a mirror that returns content directly via a WebFetch-style HTTP fetch. Once content is retrieved, the skill extracts the article body and formats it for presentation (plain text or Markdown).

When to use it

You have a Medium article URL that shows a paywall or clipped content.
The article lives on a Medium-hosted domain (towardsdatascience, betterprogramming, etc.).
You need a quick programmatic way to fetch the full article text for reading or analysis.
You want a fallback sequence when the preferred mirror is unavailable.

Best practices

Try the primary mirror first; it usually returns clean HTML suitable for extraction.
URL-encode the full article URL when constructing mirror requests (slashes -> %2F, @ -> %40).
If the primary mirror returns empty or blocked content, try secondary mirrors in order.
If HTML looks garbled, re-run the fetch with instructions to extract and format as Markdown.
Be mindful of legal and terms-of-service constraints when retrieving paywalled content.

Example use cases

User shares https://medium.com/@user/article and you need the full article text.
Bulk-processing a list of Medium-hosted article URLs for offline reading or NLP.
Automated summarization pipeline that first needs the complete article body.
Debugging why an article fetch returns truncated content and switching mirrors.

FAQ

Which mirror should I try first?

Use the mirror that returns content directly via WebFetch; it produces the cleanest HTML for extraction.

What if the mirror is blocked or returns a captcha?

Fallback to other mirrors in sequence. Some services may require a browser or present captchas and will not work via programmatic fetch.

How should I encode the Medium URL?

URL-encode the full article URL (no safe characters) when embedding it into the mirror request path.