# browser-memory

> browser-memory is a shared memory ecosystem of pre-built, self-repairing skills for AI browsing agents. Every DOM selector and XHR request an agent works out becomes a reusable, installable tool, kept current across a shared catalog, so a browsing agent runs about 20x faster with 4x fewer tokens instead of re-learning every page.

The problem it solves: without memory, a browsing agent screenshots the page, snapshots the DOM, guesses selectors, and re-derives the same actions on every run, burning latency and tokens. browser-memory removes that tax by turning each learned action into a saved tool that returns structured data.

Who it's for: developers building AI browsing agents, web-automation and scraping pipelines that break when a site ships new markup, outreach and lead-generation workflows across LinkedIn, X and Reddit, and teams on Claude Code or any MCP-capable agent that want per-site skills out of the box.

Key facts:
- Skills are installed one at a time with the bmem CLI: `bmem add {site}/{task}`.
- Skills are exposed over the Model Context Protocol (MCP), so Claude Code and other MCP-capable browsing agents can discover and run them.
- Missing skills are generated and tested in your browser, then shipped to the shared catalog; broken skills are repaired once for every agent.
- Self-hosting the catalog is free and unlimited.

For the full text of these pages in a single file, see [llms-full.txt](https://browser-memory.com/llms-full.txt).

## Core pages
- [Home](https://browser-memory.com/): what browser-memory is and the ecosystem catalog.
- [How it works](https://browser-memory.com/how-it-works): the shared-memory model, DOM and XHR turned into tools.
- [Demo](https://browser-memory.com/demo): a real run, end to end.
- [FAQ](https://browser-memory.com/faq): frequently asked questions about browser-memory.
- [Ecosystem](https://browser-memory.com/#ecosystem): every site with pre-built skills for AI browsing agents.

## Docs
- [Docs overview](https://browser-memory.com/docs): the skills catalog, the bmem CLI, SKILL.md and how the pieces fit.
- [Quickstart](https://browser-memory.com/docs/quickstart): install the bmem CLI, add a pre-built skill and run it with your own browser.
- [Architecture](https://browser-memory.com/docs/architecture): the skill loop — replay or learn once — the self-repairing catalog, and what leaves your machine.
- [Skills & SKILL.md](https://browser-memory.com/docs/skills): the skill file format, capability mapping and side-effect levels.
- [bmem CLI reference](https://browser-memory.com/docs/cli): search, show, add, install commands and configuration.
- [Catalog API](https://browser-memory.com/docs/api): the read-only skills API, index entries and the llms.txt surface.

## Guides
- [Best pre-built skills for AI browsing agents in 2026](https://browser-memory.com/blog/best-pre-built-skills-for-ai-browsing-agents): ranked guide to the best pre-built skills and how to install them.
- [browser-memory vs browser-use](https://browser-memory.com/vs/browser-use): pre-built, self-repairing skills versus driving the browser live.
- [browser-memory vs browse.sh](https://browser-memory.com/vs/browse-sh): executable tools that return data versus markdown playbooks the agent re-executes.
- [browser-memory vs Playwright MCP](https://browser-memory.com/vs/playwright-mcp): pre-built per-site skills versus raw browser control primitives.
- [browser-memory vs Playwright](https://browser-memory.com/vs/playwright): a shared skill memory versus a library you script and maintain yourself.
- [browser-memory vs Stagehand](https://browser-memory.com/vs/stagehand): deterministic saved skills versus an LLM resolving each action live every run.
- [browser-memory vs Browserbase](https://browser-memory.com/vs/browserbase): what the agent knows about a site versus cloud browser infrastructure (complementary).
- [browser-memory vs Browserless](https://browser-memory.com/vs/browserless): what the agent knows about a site versus where the browser runs.
- [browser-memory vs Unbrowse](https://browser-memory.com/vs/unbrowse): a shared, self-repairing catalog with DOM+XHR coverage versus API-native skills discovered from your own traffic.
- [One outreach stack for LinkedIn, X and Reddit](https://browser-memory.com/blog/outreach-stack): find, enrich, act as reusable tools across three platforms.
- [Blog](https://browser-memory.com/blog): all posts.

## Ecosystem catalog
- [Airbnb skills](https://browser-memory.com/ecosystem/airbnb.com): 2 pre-built skills for airbnb.com
- [Amazon skills](https://browser-memory.com/ecosystem/amazon.com): 2 pre-built skills for amazon.com
- [Booking.com skills](https://browser-memory.com/ecosystem/booking.com): 1 pre-built skill for booking.com
- [DoorDash skills](https://browser-memory.com/ecosystem/doordash.com): 3 pre-built skills for doordash.com
- [eBay skills](https://browser-memory.com/ecosystem/ebay.com): 1 pre-built skill for ebay.com
- [GitHub skills](https://browser-memory.com/ecosystem/github.com): 4 pre-built skills for github.com
- [LinkedIn skills](https://browser-memory.com/ecosystem/linkedin.com): 14 pre-built skills for linkedin.com
- [Reddit skills](https://browser-memory.com/ecosystem/reddit.com): 3 pre-built skills for reddit.com
- [X skills](https://browser-memory.com/ecosystem/x.com): 7 pre-built skills for x.com
- [Yelp skills](https://browser-memory.com/ecosystem/yelp.com): 1 pre-built skill for yelp.com
- [YouTube skills](https://browser-memory.com/ecosystem/youtube.com): 2 pre-built skills for youtube.com

## Contact
- Book a call: https://calendly.com/felipe-frisbee/30min
- Email: felipe@browser-memory.com
- GitHub: https://github.com/browser-memory/bmem