Can this skill access paid Substack content?

Yes, it includes an auth-capture script that allows you to log in manually via a browser once; your credentials are then cached to allow access to any publications you are subscribed to.

What technical requirements are there to run this?

You need Python installed, along with the Playwright library and Chromium browser, which the skill can help you set up during the first-time configuration phase.

How does the skill manage large numbers of articles without hitting token limits?

It uses an adaptive detail level system. For larger corpora (e.g., 50+ articles), it automatically switches to lower-detail theme clustering to preserve context window and budget.

What is 'Budget Mode' in the Substack Research skill?

Budget Mode limits extraction depth, caps analysis rounds, and skips non-essential file generation to minimize API costs while still delivering core research insights.

Substack Research Toolkit

Name: Substack Research Toolkit
Author: bogheorghiu

bybogheorghiu

0•

Web Scraping & Data Collection

Extracts and analyzes long-form content from Substack publications to synthesize independent journalist perspectives and expert commentary.

The Substack Research skill empowers Claude to scrape, parse, and analyze content from independent publications, providing a window into specialized discourse often missed by mainstream media. It features an automated browser authentication system to handle paywalls and 2FA, adaptive detail levels that adjust based on corpus size to optimize token usage, and sophisticated pattern analysis tools. Whether building a research corpus, mapping intellectual frameworks, or cross-referencing claims across multiple authors, this skill automates the heavy lifting of gathering and synthesizing long-form written data.

Key Features

01Multi-level parsing (0-10) to switch between quick theme clustering and deep argument mapping

02Automated content extraction with Playwright-based authentication for public and paywalled posts

030 GitHub stars

04Per-author and cross-author synthesis to identify consensus, divergence, and recurring blind spots

05Saturation detection algorithms that dynamically stop processing when new insights diminish

06Flexible export options including structured Markdown, machine-readable JSON, and relational memory updates

Use Cases

01Rapidly triaging large archives of 100+ articles to find relevant topic clusters and key arguments

02Analyzing the evolution of a specific journalist's thesis or expert perspective over time

03Triangulating independent reporting with mainstream sources to detect structural media blind spots

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add bogheorghiu/ex-cog substack-research

For use in Claude.ai and ChatGPT

Key Features

01Multi-level parsing (0-10) to switch between quick theme clustering and deep argument mapping

02Automated content extraction with Playwright-based authentication for public and paywalled posts

030 GitHub stars

04Per-author and cross-author synthesis to identify consensus, divergence, and recurring blind spots

05Saturation detection algorithms that dynamically stop processing when new insights diminish

06Flexible export options including structured Markdown, machine-readable JSON, and relational memory updates

Use Cases

01Rapidly triaging large archives of 100+ articles to find relevant topic clusters and key arguments

02Analyzing the evolution of a specific journalist's thesis or expert perspective over time

03Triangulating independent reporting with mainstream sources to detect structural media blind spots