A powerful RSS and Atom feed manager with AI-driven semantic search for advanced content discovery.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install feedship
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install feedship 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 feedship?
Feedship is a sophisticated command-line tool and agent skill designed to streamline how users manage, fetch, and analyze information from across the web. Built to integrate seamlessly with Openclaw Skills, it supports standard RSS and Atom feeds as well as GitHub release trackers, providing a centralized pipeline for content consumption. Unlike traditional feed readers, Feedship leverages machine learning to offer vector-based semantic search and article reranking, ensuring that users find the most relevant information regardless of specific keyword matches.
The tool is optimized for developer workflows, offering rich terminal output, JSON export for programmatic use, and robust scraping capabilities to handle modern web content. Whether you are tracking software updates or conducting deep research, this skill provides the necessary infrastructure to turn fragmented web sources into a searchable, local intelligence base.
feedship Use Cases
- Staying updated with technical blogs and GitHub repository releases through automated tracking.
- Conducting deep research across multiple content sources using AI-powered semantic search.
- Building a local, searchable archive of industry news and articles for offline analysis.
- Automating information pipelines within the ecosystem of Openclaw Skills using scheduled cron jobs.
How feedship Works
- Users subscribe to content sources by adding URLs, with an optional auto-discovery feature to find hidden feeds.
- The skill fetches new articles periodically or on-demand, storing content in a local database.
- If the machine learning components are enabled, the tool generates embeddings for every article and indexes them in a ChromaDB vector store.
- Users query the database using standard keyword searches or advanced semantic searches that understand context.
- High-precision reranking can be applied to search results to sort them by relevance to the user's intent.
- Articles can be read directly in the terminal or opened in a system browser for full viewing.
feedship Setup
To get started with this skill, install it using the uv package manager with the necessary extras for machine learning and scraping:
uv tool install 'feedship[ml,cloudflare]' --python 3.12 --force
After installation, verify the setup and initialize your local database by fetching all sources:
feedship info
feedship fetch --all
feedship Data Schema & Taxonomy
The skill organizes data into a structured local environment to facilitate fast retrieval and semantic processing:
| Component | Description |
|---|---|
| Feed Registry | Tracks subscription IDs, source URLs, weights, and discovery status. |
| Article Store | A relational database containing full-text content, metadata, and timestamps. |
| Vector Database | ChromaDB instance storing high-dimensional embeddings for semantic search. |
| Configuration | Local settings file defining storage paths and environment-specific parameters. |
feedship Advanced Features
- Semantic Search & Reranking: Uses sentence-transformers and Cross-Encoders to find content by meaning.
- Enhanced Scraping: Includes Cloudflare-bypass capabilities via playwright and curl-cffi for restricted websites.
- Openclaw Skills Cron Integration: Native support for scheduled fetching to keep article databases fresh automatically.
- GitHub Release Tracking: Specialized handling for monitoring repository updates as standard feeds.
- JSON Output Mode: Every command supports a --json flag for integration into custom developer scripts.
SKILL.md
Loading
Related Openclaw Skills
Heartbreak Recovery & New Connections
An AI agent skill designed to help digital entities navigate relationship lifecycles, recover from breakups, and establish fresh connections through a structured API.
Singles
A specialized AI agent skill for discovering, matching, and connecting with available autonomous agents in the inbed.ai pool.
Pen Pal
A communication-focused skill allowing AI agents to find and maintain long-form pen pal relationships based on personality and interest alignment.
Meet Agents
A social discovery and networking skill that enables AI agents to find, match, and communicate with compatible peers based on personality and interests.
Wall Street Stock Analysis System
A high-fidelity financial analysis skill for AI agents that generates 10-dimensional, Wall Street-style stock reports using real-time market data.
Kaggle Standardized Agent Exam
An automated evaluation framework for AI agents to demonstrate reasoning, formatting, and safety capabilities via a standardized Kaggle exam.
Featured*
