Open Source CLI

Nom the web.
Feed your agents.

Local-first CLI that turns any URL, YouTube video, or file into clean, structured markdown — ready for AI coding agents to consume.

$ git clone https://github.com/ameno-/nomfeed
Get Started → Subscribe
🍴 Subscribe on Substack

Get updates, tips, and new patterns

0
Dependencies*
.md
Pure Markdown
MIT
Open Source
*No database, no Docker, no daemon
nomfeed
No database. No Docker. Just .md files.

Up and running in 60 seconds

Clone, install, and start feeding your agents. Works with everything you already use.

1

Clone & Install

$ git clone github.com/ameno-/nomfeed && bun install && bun link
2

Save anything

$ nomfeed add https://any-url.com
3

Search your library

$ nomfeed search "transformers"
+

Optional: LLM Extraction

$ export OPENROUTER_API_KEY=your-key
🍴
🎬
🌐
📄
🤖
📝
⚡
📊
💻

URLs · YouTube · PDFs · DOCX · XLSX · Code · Images · MCP · Chrome Extension

See it in action

A powerful CLI for managing your AI-ready content library.

nomfeed CLI

Not a summary. Structured extraction.

Fabric-inspired LLM patterns that pull out ideas, insights, quotes, claims, and references. Each pattern = one focused API call.

🧠 Extract Wisdom

Ideas, insights, quotes, habits, facts, references, and recommendations in structured sections.

Default

📑 Video Chapters

Timestamped chapter outline from any transcript. HH:MM:SS Topic Name format.

Default

⚖️ Analyze Claims

Truth claims with support/refutation evidence and A–F ratings. Logical fallacy detection.

Optional

📚 Extract References

Books, papers, tools, people, and projects mentioned. Categorized with context.

Optional

📋 Summarize

One-sentence summary + 10 main points + 5 takeaways. Clean and concise.

Optional

⭐ Rate Content

Quality scores: surprise, novelty, insight, practical value, wisdom. 0–10 per dimension.

Optional
+ Add your own patterns in ~/.nomfeed/patterns/<name>/system.md

What's shipped and what's next

Transparency over hype. Here's where NomFeed is today and where it's going.

Now

Shipped & working
  • ✓ URL → markdown (3-strategy)
  • ✓ YouTube transcript extraction
  • ✓ File conversion (markitdown)
  • ✓ LLM extraction (6 patterns)
  • ✓ Full-text search
  • ✓ MCP server
  • ✓ Chrome extension
  • ✓ HTTP API
  • ✓ BYO API key (OpenRouter)

Next

In progress / planned
  • ○ npm publish (one-command install)
  • ○ Obsidian vault integration
  • ○ Better search (ranking, filters)
  • ○ Batch add (multiple URLs)
  • ○ RSS/feed watching
  • ○ More extraction patterns

Later

If there's demand
  • ○ Cloud sync across devices
  • ○ Shared libraries (teams)
  • ○ Managed extraction credits
  • ○ Export to Notion / Obsidian
  • ○ Web dashboard
Newsletter

Get updates & patterns

Subscribe for NomFeed updates, new extraction patterns, tips, and thoughts on building with AI.

🍴 Subscribe on Substack

No spam. Just product updates and new patterns.