feedship

Personal information system for collecting, subscribing to, and organizing information sources from the internet.

Features

• Feed subscription - RSS/Atom feeds and GitHub releases
• Web article extraction - Crawl webpages with Readability
• Full-text search - FTS5-powered search across all content
• Semantic search - Vector embeddings with ChromaDB
• CLI tool - Full-featured command-line interface

Tech Stack

Python 3.10+ | click | feedparser | scrapling | sqlite3 | ChromaDB | sentence-transformers

Installation

pip install feedship
# or
uv pip install feedship

Optional Dependencies

# ML/AI features for auto-tagging
pip install feedship[ml]

# All features
pip install feedship[ml]

Quick Start

Add a Feed

feedship feed add <url> [options]

# Examples:
feedship feed add https://example.com/feed.xml
feedship feed add https://github.com/python/cpython

# Options:
--discover [on|off]          Enable feed discovery (default: on)
--automatic [on|off]          Auto-add all discovered feeds (default: off)
--discover-depth N           Discovery depth 1-10 (default: 1)
--weight FLOAT               Feed weight for semantic search (default: 0.3)

Fetch & List

feedship fetch --all        # Fetch all feeds
feedship feed list          # List all feeds
feedship article list        # List articles
feedship article list --limit 50

Search

feedship search "machine learning"
feedship search "python" --limit 10

Documentation

• @docs/feed.md - Feed provider architecture, fetch flow, refactoring status
• @docs/providers.md - Provider/TagParser interfaces, registration
• @docs/structure.md - Application structure, source files, structural rules
• @docs/cli-commands.md - CLI command reference