Glossary

What is Browsing Context?

Concept By David Hamilton
Definition

Browsing Context is the collection of web pages, bookmarks, saves, and browsing signals that represent what you have found valuable online over time.

What browsing context means

Browsing context is a simple idea: it is the sum of what you have found valuable online. Every article you bookmarked, every tweet you liked, every Reddit post you saved, every LinkedIn article you kept for later. Taken together, these signals paint a picture of your interests, your research, and your professional focus.

Most people do not think of their saved content this way. Bookmarks feel like a utility, a way to find a page again. But viewed as context, they become something more powerful: a personalised knowledge base that reflects months or years of curation.

Why browsing context matters for AI

AI assistants like Claude are trained on broad knowledge but know nothing about you personally. They cannot tell you what articles you saved last week about a topic, what thread changed your thinking on a subject, or which resource you keep coming back to.

When you give an AI access to your browsing context, this changes. Instead of searching your memory for “that article about database indexing I saved somewhere”, you ask Claude and it searches your actual saves. Instead of starting research from scratch, the AI can build on what you have already collected.

This is the core idea behind semantic bookmarking: saved content becomes searchable context rather than a static archive. ContextBolt makes this possible by collecting your bookmarks and social saves and exposing them as searchable data through MCP.

The retrieval problem

The reason browsing context is underused is simple: retrieval is broken. People save content across Twitter/X, Reddit, LinkedIn, browser bookmarks, and read-it-later apps. Each platform has its own search (if it has search at all), and none of them talk to each other.

The result is that most browsing context is effectively lost. Studies suggest the average person has hundreds or thousands of saved items they can no longer find. The act of saving creates a false sense of security: you feel like you have stored the information, but in practice it is buried.

ContextBolt solves this by pulling bookmarks and saves into a single searchable layer. Combined with topic clustering, your browsing context becomes organised and accessible, both to you and to AI assistants.

Browsing context in practice

With ContextBolt connected to Claude Desktop or Cursor, your browsing context becomes part of your AI workflow:

The value compounds over time. The more you save, the richer your browsing context becomes, and the more useful AI access to it gets.

Related terms

Frequently asked questions

What counts as browsing context? +
Browsing context includes anything you've intentionally saved or engaged with online: bookmarks, Twitter/X likes, Reddit saves, LinkedIn saved posts, read-it-later items, and highlighted passages. It can also include browsing history, though saved content is typically more meaningful because you actively chose to keep it.
Why is browsing context useful for AI? +
AI assistants are powerful but generic. They do not know what you have read, saved, or care about. Giving an AI access to your browsing context turns it from a general-purpose tool into a personalised assistant that can reference your actual research, interests, and saved knowledge.
How is browsing context different from browsing history? +
Browsing history is every page you visited, including accidental clicks and quick bounces. Browsing context is the curated subset: content you actively chose to save, bookmark, or engage with. It is a much stronger signal of what you actually find valuable.
Is sharing my browsing context with AI safe? +
With ContextBolt, your browsing context stays on your device. The MCP server runs locally, and AI assistants access your data through a local connection. Your bookmarks and saves are never uploaded to external servers unless you explicitly choose to use a remote connection.