Kosu for Openclaw

Kosu is a mindful content queue designed for AI agents to filter, curate, and manage reading lists with an intentional decay mechanic.

matthewspear
v1.1.0
Mar 13, 2026
0
928
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install kosu

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install kosu using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is Kosu?

Kosu (Japanese for to strain or filter) is a purpose-built content management system that rejects the traditional read-it-later graveyard. It operates on a philosophy of intentionality where items automatically decay and archive if not engaged with within a set timeframe. This unique mechanic ensures your queue remains relevant and manageable rather than becoming an infinite pile of unread tabs.

As one of the premier agent-native Openclaw Skills, Kosu provides a robust API-first interface. It is specifically designed for AI agents to act as curators—suggesting relevant content, reordering priorities, and tracking consumption—while the human user focuses solely on high-value reading and learning.

Kosu Use Cases

  • Automatically cleaning up unread articles through a 14-day stale warning and 28-day auto-archive.
  • Delegating content curation to an AI agent to suggest URLs based on specific research interests.
  • Managing high-priority reading lists by programmatically reordering items to the top of the queue.
  • Syncing YouTube watch lists with auto-enriched metadata like channel name and duration.
  • Preventing information hoarding by enforcing queue limits on the free tier.

How Kosu Works

  1. The AI agent connects to the Kosu API using a secure KOSU_API_KEY.
  2. When a URL is added, the server automatically enriches the entry with metadata such as titles, thumbnails, and duration.
  3. Items are placed in a queue where a decay timer begins; users or agents can extend this timer by marking an item as opened.
  4. The agent manages the lifecycle of the content by moving items between states: queue, read, archived, or suggested.
  5. If content remains untouched past the decay threshold, Kosu automatically moves it to the archive to maintain a clean workspace.

Kosu Setup

For Openclaw Skills Agents

Create the directory and download the skill definition:

mkdir -p ~/.openclaw/skills/kosu
curl -s https://usekosu.com/skill.md > ~/.openclaw/skills/kosu/SKILL.md

Authentication

  1. Sign up at usekosu.com.
  2. Generate an API key in Settings.
  3. Export the key to your environment:
export KOSU_API_KEY='your_api_key_here'

Kosu Data Schema & Taxonomy

Kosu utilizes a structured schema for content items, accessible via the [https://usekosu.com/api/v1](https://usekosu.com/api/v1) base URL.

Field Type Description
id String Unique identifier (prefix itm_)
url String The canonical link to the content
source Enum Classification (youtube, article, book, etc.)
state Enum Current status (queue, read, archived, deleted)
duration_minutes Integer Estimated time to consume
opened_at Timestamp Last time the decay protection was extended

Kosu Advanced Features

  • Server-side deduplication based on canonical URLs to prevent duplicate entries.
  • Granular position management allowing agents to move items before or after specific IDs.
  • Suggestion engine where agents can propose content with a specific reason field for user approval.
  • Automatic YouTube enrichment providing channel names and video durations without external API calls.
  • Cursor-based and offset-based pagination for high-performance queue syncing.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*