Skip to content

First Steps

Start the Dashboard

km serve

Open http://localhost:3749 in your browser. You'll see the Knowmarks dashboard with onboarding cards guiding you through setup.

Save Your First Bookmark

From the CLI:

km save https://example.com --note "Interesting article about X"

Or from the dashboard: use the unified input bar at the bottom of the screen. Paste a URL and press Enter.

Knowmarks fetches the page, extracts content, generates an embedding for semantic search, classifies the content type, and adds it to your collection.

Search Your Collection

km find "machine learning best practices"

Results are ranked by hybrid search — a combination of semantic similarity (what you meant) and keyword matching (what you typed).

Type your query in the input bar. Results appear instantly with relevance explanations showing why each item matched.

Prefix your query with a question word or phrase to get a generative answer grounded in your saved content:

  • "What are the best practices for error handling in Go?"
  • "Summarize what I've saved about React server components"

Tip

Conversational search requires an LLM provider to be configured. Without one, queries fall back to standard hybrid search.

Understand Pulse

Click Pulse in the sidebar (or run km status from the CLI) to see your collection health at a glance:

  • Recent additions — What you've saved lately
  • Stale items — Content that may need attention (dead links, outdated)
  • Emerging topics — Clusters that are growing with new content
  • Connector health — Status of your import connectors
  • Project activity — Which projects have new relevant items

Pulse is your daily starting point for knowledge governance.