Skip to main content
Glama

Instapaper MCP Server

A comprehensive Model Context Protocol (MCP) server for Instapaper integration. This server allows Claude and other MCP clients to interact with your Instapaper account - reading, saving, organizing, and analyzing your articles.

Like what you see?

Buy me a coffee: https://cmdzed.com/#/portal/support

Features

šŸ› ļø Tools (Actions)

Content Management:

  • add_bookmark - Save articles with title, description, and folder

  • add_private_bookmark - Save private content without URLs (emails, notes, generated content)

  • delete_bookmark - Remove articles

  • archive_bookmark - Move articles to archive

  • unarchive_bookmark - Restore from archive

  • star_bookmark - Mark as important

  • unstar_bookmark - Remove star

  • move_bookmark - Organize into folders

  • update_read_progress - Track reading progress

Bulk Operations (Parallel Processing):

  • move_bookmarks_bulk - Move multiple bookmarks to a folder at once

  • star_bookmarks_bulk - Star multiple bookmarks in parallel

  • unstar_bookmarks_bulk - Remove stars from multiple bookmarks

  • archive_bookmarks_bulk - Archive multiple bookmarks at once

  • unarchive_bookmarks_bulk - Restore multiple bookmarks from archive

  • update_read_progress_bulk - Update reading progress for multiple articles

Folder Management:

  • list_folders - View all folders

  • create_folder - Create new folders

  • delete_folder - Remove folders

  • reorder_folders - Customize folder order

Highlights:

  • add_highlight - Save important passages

  • list_highlights - View highlights for an article

  • delete_highlight - Remove highlights

Search & Discovery:

  • list_bookmarks - List articles from folders with sync support

  • search_bookmarks - Find articles by title, URL, or description

Content Access:

  • get_article_content - Fetch full text of a single article

  • get_articles_content_bulk - Fetch content from multiple articles at once for bulk analysis

šŸ“š Resources (Data Claude Can Read)

  • instapaper://bookmarks/unread - All unread articles

  • instapaper://bookmarks/archive - Archived articles

  • instapaper://bookmarks/starred - Starred articles

  • instapaper://folders - List of folders

  • instapaper://folder/{folder_id} - Articles in a specific folder

  • instapaper://article/{bookmark_id} - Full text of an article

šŸ’” Prompts (Reusable Workflows)

  • weekly_reading_digest - Organized summary of unread articles

  • recommend_next_read - AI-powered reading suggestions

  • research_synthesis - Synthesize insights from articles on a topic

  • organize_backlog - Suggest folder organization

  • archive_candidates - Identify old articles to archive

  • save_as_private_bookmark - Guidelines for saving private content and generated insights

Related MCP server: raindrop-mcp-server

Installation

Prerequisites

  1. Node.js 18+ installed on your system

  2. Instapaper account with API credentials

Step 1: Get Instapaper API Credentials

  1. Go to https://www.instapaper.com/api

  2. Fill out the API access request form

  3. Wait for approval (usually takes a few days)

  4. You'll receive a Consumer Key and Consumer Secret

Step 2: Clone and Install

# Clone or download this repository
cd instapaper-mcp-server

# Install dependencies
npm install

# Create your environment file
cp .env.example .env

Step 3: Configure Environment Variables

Edit .env and add your credentials:

INSTAPAPER_CONSUMER_KEY=your_consumer_key_here
INSTAPAPER_CONSUMER_SECRET=your_consumer_secret_here
INSTAPAPER_USERNAME=your_instapaper_email@example.com
INSTAPAPER_PASSWORD=your_instapaper_password

Step 4: Build the Server

npm run build

Usage with Claude Desktop

Configure Claude Desktop

Add this to your Claude Desktop configuration file:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "instapaper": {
      "command": "node",
      "args": ["/absolute/path/to/instapaper-mcp-server/build/index.js"],
      "env": {
        "INSTAPAPER_CONSUMER_KEY": "your_consumer_key",
        "INSTAPAPER_CONSUMER_SECRET": "your_consumer_secret",
        "INSTAPAPER_USERNAME": "your_email",
        "INSTAPAPER_PASSWORD": "your_password"
      }
    }
  }
}

Important: Replace /absolute/path/to/instapaper-mcp-server with the actual full path.

Restart Claude Desktop

After updating the config, restart Claude Desktop completely (quit and reopen).

Example Usage

Once configured, you can interact with Instapaper through Claude. Here are real-world examples showing what happens when you make requests:

Example 1: Organize Recent Bookmarks by Topic

You ask Claude:

"Organize my recent bookmarks by topic"

Claude's workflow:

  1. Fetches your unread bookmarks using list_bookmarks

  2. Analyzes the titles and URLs to identify topics (e.g., "AI/ML", "Design", "Product Management")

  3. Creates new folders for each topic using create_folder

  4. Gets the bookmark IDs for articles in each category

  5. Uses move_bookmarks_bulk to organize them all at once

You see:

āœ“ Fetched 47 unread bookmarks
āœ“ Identified topics: AI & Machine Learning (12), UX Design (15), Product Strategy (11), Other (9)
āœ“ Created folders: "AI & ML", "UX Design", "Product Strategy"
āœ“ Moved 38 bookmarks to topic folders:
  - AI & ML: 12 moved
  - UX Design: 15 moved  
  - Product Strategy: 11 moved

Your bookmarks are now organized by topic!

Example 2: Review Top Articles and Sync Across Devices

You ask Claude:

"Show me my starred articles and mark the most interesting ones as fully read"

Claude's workflow:

  1. Fetches starred bookmarks using list_bookmarks with folder: "starred"

  2. Retrieves full article text for your starred items using get_articles_content_bulk

  3. Analyzes the content and identifies the most compelling articles

  4. Updates reading progress for selected articles using update_read_progress_bulk

You see:

āœ“ Found 24 starred articles
āœ“ Fetched full text from 24 articles in parallel
āœ“ Most interesting articles:
  1. "The Future of Design Systems" - Design patterns at scale
  2. "Rethinking Product Strategy" - How great teams think about roadmaps
  3. "AI-Assisted Development" - New tools reshaping engineering

āœ“ Marked these 3 articles as fully read
āœ“ Progress updated across your devices

Example 3: Batch Process and Archive

You ask Claude:

"Archive all articles older than 3 months that I haven't starred"

Claude's workflow:

  1. Lists all unread bookmarks with timestamps

  2. Filters for articles older than 3 months

  3. Excludes any that are starred

  4. Uses archive_bookmarks_bulk to move them all to archive

You see:

āœ“ Analyzed 128 bookmarks
āœ“ Found 47 articles older than 3 months, not starred
āœ“ Archived 47 bookmarks in parallel:
  - Successfully archived: 47
  - Errors: 0

Your queue is now more focused on recent content!

Example 4: Research Synthesis

You ask Claude:

"Find all my articles about remote work culture and create a summary"

Claude's workflow:

  1. Searches bookmarks using search_bookmarks for "remote work culture"

  2. Fetches full article content using get_articles_content_bulk

  3. Synthesizes key insights from all articles

  4. Optionally stars the most relevant articles using star_bookmarks_bulk

You see:

āœ“ Found 8 articles about remote work culture
āœ“ Retrieved full text from all 8 articles
āœ“ Analysis: Key themes across your reading:
  - Asynchronous communication is critical (5/8 articles)
  - Building trust in distributed teams (6/8 articles)
  - Time zone management strategies (4/8 articles)

āœ“ Key takeaways:
  - Documentation and clarity reduce misunderstandings
  - Regular 1-on-1s maintain relationships
  - Async-first mindset improves flexibility

āœ“ Starred your 3 most-cited articles for reference

Example 5: Quick Content Pipeline

You ask Claude:

"Save this research summary to Instapaper with the 'AI Research' folder"

Claude's workflow:

  1. Saves article using add_bookmark to specified folder

  2. Confirms save is complete

You see:

āœ“ Saved to Instapaper:
  - Title: "AI Research Summary - Week 48"
  - Folder: AI Research
  - Ready to read on any device

The article will sync to your Kindle in a few minutes.

Development

Run in Development Mode

# Watch for changes
npm run watch

# Test with MCP Inspector
npm run inspector

Project Structure

instapaper-mcp-server/
ā”œā”€ā”€ src/
│   ā”œā”€ā”€ index.ts              # Main MCP server
│   └── instapaper-client.ts  # Instapaper API client
ā”œā”€ā”€ build/                     # Compiled JavaScript
ā”œā”€ā”€ .env                       # Your credentials (git-ignored)
ā”œā”€ā”€ .env.example              # Template
ā”œā”€ā”€ package.json
└── tsconfig.json

Troubleshooting

Authentication Errors

If you see authentication errors:

  1. Verify your credentials in .env

  2. Ensure you have API access (check your Instapaper email)

  3. Try re-authenticating by restarting Claude Desktop

Tools Not Showing Up

  1. Check that the path in claude_desktop_config.json is absolute

  2. Verify the server builds successfully: npm run build

  3. Check Claude Desktop logs for errors

API Rate Limits

Instapaper has rate limits. If you hit them:

  • Wait a few minutes before trying again

  • Reduce the frequency of requests

  • Use resources (reading data) instead of tools when possible

Advanced Features

Reading Full Article Text

"Read the full text of bookmark ID 12345 and summarize it"

Claude will use the instapaper://article/{bookmark_id} resource to access the full article content.

Folder-Based Research

"Analyze all articles in my 'User Research' folder and identify common themes"

Highlight Management

"Add a highlight to bookmark 12345: 'Users prefer familiar patterns over novel ones' at position 1000"

"Show me all my highlights from the article about design systems"

Bulk Operations

Bulk Content Analysis:

"Get the content from bookmarks 123, 124, and 125 and identify common themes"

"Fetch all articles in my 'AI Research' folder and analyze them for recent developments"

Bulk Organization:

"Move bookmarks 100, 101, 102, and 103 to my 'UX Research' folder"

"Star all the articles I just found about design systems"

"Archive all bookmarks 200-210"

Bulk Status Updates:

"Mark bookmarks 50, 51, and 52 as fully read (progress 1.0)"

"Unarchive my last 10 archived articles"

All bulk operations execute requests in parallel for maximum efficiency and return detailed results including success count, failure count, and per-item status.

License

MIT

Contributing

Contributions welcome! Please feel free to submit issues or pull requests.

Resources

Changelog

v1.2.0 (Latest)

  • Added 5 new bulk operation tools with parallel processing:

    • star_bookmarks_bulk - Star multiple bookmarks at once

    • unstar_bookmarks_bulk - Remove stars from multiple bookmarks

    • archive_bookmarks_bulk - Archive multiple bookmarks in parallel

    • unarchive_bookmarks_bulk - Restore multiple bookmarks from archive

    • update_read_progress_bulk - Update progress for multiple articles

  • Added list_bookmarks tool with advanced filtering and sync support

  • Added reorder_folders tool for custom folder organization

  • Enhanced bulk operation documentation with examples

  • All bulk tools provide detailed success/failure statistics

v1.1.0

  • Added get_article_content tool for fetching individual article text

  • Added get_articles_content_bulk tool for bulk article content retrieval

  • Enhanced parallel processing for better performance

  • Improved error handling in bulk operations

v1.0.0

  • Initial release

  • Complete implementation of Instapaper API

  • All tools, resources, and prompts

  • OAuth 1.0 authentication

  • Comprehensive error handling

Available Tools

26 tools
add_bookmarkA

Save an article or web page to Instapaper. Use this when the user wants to save a URL for later reading. Returns the bookmark ID, title, and URL. You can optionally provide a title, description/notes, and a folder ID to organize the article.

Inputs: url (required), title, folder_id, description.

add_highlightB

Add a highlight (excerpt) to an article. Specify the text to highlight and its position in the article. Use this to mark important passages, quotes, or key insights for later reference.

Inputs: text (required), position (required), bookmark_id (required).

add_private_bookmarkA

Save private content without a URL to Instapaper (e.g., from emails, notebooks, clipped text, Slack messages). Private bookmarks are not shared and don't have URLs. Perfect for personal notes, clipped content, or information from sources without URLs. Requires HTML content to be provided.

Inputs: content (required), source_label (required), title, folder_id, +1 more.

archive_bookmarkA

Move a bookmark to the archive. Archiving removes articles from your unread queue without deleting them, useful for marking articles as done or temporarily hiding them. The bookmark can be unarchived later.

Inputs: bookmark_id (required).

archive_bookmarks_bulkA

Move multiple bookmarks to archive in parallel. Use this to quickly mark a set of articles as done or hide them from your unread queue. Returns success count and any failures.

Inputs: bookmark_ids (required).

create_folderC

Create a new folder for organizing bookmarks. Provide a folder name/title. Folders help categorize articles by topic, project, or any organizational system you prefer.

Inputs: title (required).

delete_bookmarkA

Permanently remove a bookmark from Instapaper. This action cannot be undone. Use this when the user wants to delete an article they no longer need.

Inputs: bookmark_id (required).

delete_folderA

Remove a folder from your Instapaper account. When a folder is deleted, all bookmarks in it are automatically moved to the unread section. Use this to clean up unused folders.

Inputs: folder_id (required).

delete_highlightC

Remove a specific highlight from an article. Use this to clean up highlights or remove excerpts you no longer need.

Inputs: highlight_id (required).

get_article_contentA

Retrieve the complete text content of a single article. Use this to analyze, summarize, or process the full article text. Requires a bookmark ID. Returns the article as plain text.

Inputs: bookmark_id (required).

get_articles_content_bulkB

Fetch the complete text of multiple articles simultaneously for efficient bulk analysis. Provide an array of bookmark IDs. Articles are retrieved in parallel for best performance. Each article is returned with its ID, content, or error status. Ideal for synthesizing information across multiple sources, comparing perspectives, or comprehensive research analysis.

Inputs: bookmark_ids (required).

list_bookmarksB

List bookmarks from a specific folder or search results. Retrieve unread, archived, or starred articles with optional limit. This is the primary way to browse your reading list. Supports filtering by folder, limit, and sync parameters for efficient bulk data retrieval.

Inputs: have, limit, folder.

list_foldersB

Retrieve all folders in your Instapaper account. This shows custom folders you have created for organizing articles. Returns folder names, IDs, and positions.

Inputs: no named parameters.

list_highlightsA

Retrieve all highlights/excerpts saved from a specific article. Shows all the important passages you have highlighted in that bookmark, useful for reviewing key points and important takeaways.

Inputs: bookmark_id (required).

move_bookmarkB

Organize a bookmark by moving it to a specific folder. Use this to categorize articles (e.g., move research articles to a "UX Research" folder). Requires both the bookmark ID and target folder ID.

Inputs: folder_id (required), bookmark_id (required).

move_bookmarks_bulkA

Move multiple bookmarks to the same folder in parallel. While Instapaper API only supports single moves, this tool batches them efficiently using parallel requests. Ideal for organizing large sets of articles after search or bulk discovery. Returns success count and any failures.

Inputs: folder_id (required), bookmark_ids (required).

reorder_foldersC

Reorder your folders in a custom sequence. Provide an array of folder IDs with their desired positions. This helps organize your folder structure for better workflow.

Inputs: folder_order (required).

search_bookmarksB

Find bookmarks by searching through titles, URLs, and descriptions. Supports optional filtering by folder (use "unread", "archive", "starred", or a specific folder_id) and result limits. Useful for discovering articles on specific topics or finding previously saved content.

Inputs: query (required), limit, folder.

star_bookmarkA

Mark a bookmark as starred/important. Starring articles highlights them in your Instapaper library and helps you track your favorite or most valuable reads.

Inputs: bookmark_id (required).

star_bookmarks_bulkB

Mark multiple bookmarks as starred in parallel. Use this to quickly highlight a set of important articles you want to track or prioritize. Returns success count and any failures.

Inputs: bookmark_ids (required).

unarchive_bookmarkA

Restore a bookmark from the archive back to your unread list. Use this to bring back articles you want to read or reconsider.

Inputs: bookmark_id (required).

unarchive_bookmarks_bulkA

Restore multiple bookmarks from archive in parallel. Use this to bring back a group of articles to your unread list. Returns success count and any failures.

Inputs: bookmark_ids (required).

unstar_bookmarkA

Remove the star from a bookmark. Use this when an article is no longer a favorite or needs to be deprioritized in your library.

Inputs: bookmark_id (required).

unstar_bookmarks_bulkA

Remove stars from multiple bookmarks in parallel. Use this to deprioritize a group of articles at once. Returns success count and any failures.

Inputs: bookmark_ids (required).

update_read_progressB

Track reading progress on a bookmark. Provide a progress value between 0.0 (not started) and 1.0 (finished). Useful for tracking where you left off in long articles or research papers.

Inputs: progress (required), bookmark_id (required).

update_read_progress_bulkA

Update reading progress for multiple bookmarks in parallel. Useful for batch-updating progress on a set of related articles or bulk-marking articles as read. Progress ranges from 0.0 (not started) to 1.0 (finished). Returns success count and any failures.

Inputs: progress (required), bookmark_ids (required).

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    B3.4/5.0

    Scored across 26 tools

    Disambiguation4/5

    Most tools target clearly distinct resource+action pairs (folders, highlights, article content, bookmark state changes). The main risk is the five single/bulk pairs (star_bookmark vs star_bookmarks_bulk, archive/bookmark vs archive_bookmarks_bulk, etc.), which an agent could misselect, though the descriptions clearly mark the plural batch variants. list_bookmarks vs search_bookmarks also overlap slightly but are adequately differentiated.

    Naming Consistency4/5

    Nearly all tools use consistent snake_case verb_noun patterns (list_folders, add_highlight, archive_bookmark). Minor deviations: bulk suffix placement and pluralization vary (star_bookmarks_bulk vs star_bookmark, update_read_progress_bulk, get_articles_content_bulk), and add_private_bookmark mixes a qualifier into the noun. Still readable and predictable overall.

    Tool Count3/5

    26 tools is heavy for what is essentially bookmark management, largely because nearly every bookmark action has both a single and a bulk variant (10 of 26 are bulk tools). The bulk tools are defensible for efficiency, but the surface feels inflated relative to the domain.

    Completeness4/5

    Bookmark lifecycle is thoroughly covered: create (URL and private HTML), delete, browse/search, fetch content, move, archive/unarchive, star/unstar, read progress, plus highlights and folders. Gaps exist: no update/edit bookmark metadata (title, description) and no rename_folder or update_folder, so the folder surface is only partially complete.

    Maintenance

    ActivityInactive
    ResponsivenessLimited data

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      A
      quality
      D
      maintenance
      Enables Claude to interact with the Readwise Reader API, allowing for saving, listing, updating, and deleting documents with complete metadata and content access through natural language.
      6
      47
      MIT
    • A
      license
      Not graded
      quality
      C
      maintenance
      Enables Claude on web, desktop and mobile to read, mark and manage Inoreader feeds through a self-hosted remote MCP connector, exposing ten tools for unread counts, subscriptions, folders, article retrieval, starring, subscribing and marking articles read or unread.
      MIT