Weekly Content Digest

Summarizes your saved articles, bookmarked videos, newsletter archive, and reading list backlog into a prioritized weekly digest. Integrates with Pocket, Instapaper, Raindrop.io, YouTube Watch Later, browser bookmarks, and email newsletters to surface the content most worth your limited reading time, with AI-generated summaries so you can decide what deserves a full read.

Supported Platforms & Integrations

Platform	Setup Method	Auth Type	Notes
Pocket	Pocket API	OAuth 2.0	Reads saved articles with tags and favorites
Instapaper	Instapaper API	OAuth 1.0a	Reads bookmarks and highlights
Raindrop.io	Raindrop API	OAuth 2.0	Reads collections and bookmarks
YouTube	YouTube Data API v3	OAuth 2.0	Watch Later playlist and subscriptions

When to Use This Skill

• Use this when your reading list grows faster than you can read
• Use this when you save 20+ articles per week but read fewer than 5
• Use this when you want AI-generated summaries to triage what to actually read
• Consider alternatives when you prefer full-text reading without summaries (use Pocket directly)

Quick Start

Copy
# Minimal configuration - content-digest.yml
skill: weekly-content-digest
schedule: "saturday-09:00"
sources:
  pocket:
    enabled: true
    consumer_key_env: "POCKET_CONSUMER_KEY"
    access_token_env: "POCKET_ACCESS_TOKEN"
    include_tags: []
    favorites_first: true
  youtube:
    enabled: true
    playlist: "Watch Later"
    max_videos: 10
  newsletters:

Copy
claude /weekly-content-digest

Expected Output

WEEKLY CONTENT DIGEST - March 15, 2026
34 items saved this week | 15 selected for your 60-min budget

MUST-READ (3 articles, ~20 min total):

1. "The Architecture of Production AI Agents" - saved from Pocket
   Source: martinfowler.com | 12 min read | Saved: March 11
   Summary: Fowler outlines a reference architecture for production
   AI agents with supervision layers, fallback chains, and human-in-
   the-loop patterns. Argues that agent reliability requires the
   same rigor as distributed systems design.
   Why prioritized: Matches your interests (AI, architecture), from
   a high-signal source, 34 Pocket saves from your network.

...

Advanced Configuration

Platform-Specific Setup

Pocket

Copy
sources:
  pocket:
    enabled: true
    consumer_key_env: "POCKET_CONSUMER_KEY"
    access_token_env: "POCKET_ACCESS_TOKEN"
    include_tags: ["must-read", "work", "learning"]
    exclude_tags: ["archived", "reference"]
    favorites_first: true

Full Options Reference

Parameter	Type	Default	Description
schedule	string	"saturday-09:00"	Weekly digest delivery time
sources	object	required	Platform source configurations
max_items	number	15	Maximum items in the digest
summary_length	string	"medium"	Summary detail: short, medium, long
reading_time_budget	number	60	Total reading time budget in minutes
prioritization	string	"relevance"	Sort: relevance, recency, source-quality

Core Concepts

Concept	Purpose	How It Works
Time-Budget Prioritization	Fits reading to available time	Selects content that fits your stated time budget, prioritized by relevance
Cross-Source Dedup	Prevents seeing same article twice	Matches content saved in Pocket and bookmarked in browser, keeps one instance
Content Scoring	Ranks by personal relevance	Combines topic match, source reputation, save recency, and social signals
Newsletter Parsing	Extracts value from email newsletters	Identifies top links and key paragraphs from newsletter HTML, skips boilerplate

Architecture

Pocket API ──────> Article Fetcher ───┐
  Instapaper API ──> Bookmark Reader ──┤
  Raindrop API ────> Collection Reader ─┼──> Deduplicator ──> Content Scorer ──> Budget Optimizer

Workflow Examples

Scenario 1: Weekend reading plan for a busy professional

Input: 47 items saved across Pocket, YouTube, and newsletters this week. Budget: 60 minutes. Processing: Scores all items by relevance, deduplicates 5 cross-platform saves, generates summaries for top 15, calculates cumulative reading time, and selects items fitting the 60-minute budget. Output: Prioritized reading plan with "Must-Read" (3 items, 20 min), "Worth Your Time" (5 items, 25 min), and "Videos" (3 items, 15 min). Remaining 29 items deferred.

Scenario 2: Catching up after vacation

Input: 200+ unread items across all sources accumulated over 2 weeks Processing: Switches to "catch-up mode" with aggressive summarization. Groups items by topic, provides topic-level summaries instead of individual, and identifies the 5 highest-value items from the entire backlog. Output: Condensed catch-up digest with "Top 5 Must-Reads from the Last 2 Weeks" and topic summaries covering everything else in bulk.

Best Practices

1. Set a realistic reading time budget -- If you honestly have 45 minutes of reading time on weekends, set budget to 45. The skill selects the highest-value content for.

2. Use tags in Pocket/Raindrop for priority signals -- Tagging items as "must-read" when saving tells the digest to prioritize them. This takes 1 second while saving and.

3. Review video summaries before watching -- YouTube videos are time-expensive. The transcript summary lets you decide in 30 seconds whether a 15-minute video is worth.

Common Issues

1. Pocket API "invalid consumer key" error -- Ensure you registered an application at getpocket.com/developer and completed the OAuth flow. The consumer key alone is.

2. YouTube Watch Later is empty despite having videos -- Google requires YouTube Data API to be enabled in your Google Cloud Console and the OAuth scope must include.

Privacy & Data Handling

All content fetching and summarization occurs locally. Article text is fetched from original source URLs for summarization and never stored permanently. Summaries are cached for 7 days in ~/.claude/cache/content-digest/ to avoid re-fetching. Your reading list, Pocket bookmarks, and YouTube watch history are accessed through your authenticated API credentials and are not shared with any third party.