Add a ready-made recipe for scraping LinkedIn profiles into your agent.
Install every skill for a site like DoorDash or Airbnb at once.
Search a shared catalog for existing automation recipes before building your own.
Preview a skill's instructions before installing it into your agent.
| browser-memory/bmem | 1304674612/agentbench | aerdelan/housand-domaintoolmatrix | |
|---|---|---|---|
| Stars | 22 | 22 | 22 |
| Language | TypeScript | TypeScript | TypeScript |
| Setup difficulty | easy | moderate | easy |
| Complexity | 2/5 | 3/5 | 2/5 |
| Audience | developer | developer | developer |
Figures from each repo's GitHub metadata at analysis time.
Requires Node.js 20 or newer with npx available.
bmem is a command line tool that gives AI agents access to a shared catalog of reusable web skills. A skill here means a recipe for doing something on a specific website, like pulling a profile from LinkedIn or a listing from Airbnb. Instead of teaching an agent how to interact with a site from scratch every time, you can search the catalog, find a skill someone already built and tested, and add it to your agent in one step. The tool itself never opens a browser or touches your cookies. It only searches the catalog and registers a skill into your agent's own skill system. When you actually run the skill, your agent uses whatever browser tool it already has, whether that is Playwright, Puppeteer, or a browser extension it controls. This means your login sessions and cookies stay exactly where they are, since bmem is only handling the recipe, not the browsing. Using it is simple: install the CLI with npm, run a one-time setup command to connect it to your agent, then add skills either one at a time or all at once for a whole site. For example, you can add every LinkedIn skill in the catalog with a single command. The catalog already covers popular sites including LinkedIn, X, GitHub, DoorDash, Reddit, Airbnb, and Amazon, with more skills for each site than a single line suggests. Other commands let you search the catalog by keyword or site, list everything available, preview a skill's instructions before installing it, and update installed skills when the catalog changes. The project requires Node.js version 20 or newer. It is released under the MIT license, so you are free to use, modify, and redistribute it, including for commercial purposes.
A CLI that lets your AI agent search a shared catalog of website recipes and add them, running with whatever browser the agent already has.
Mainly TypeScript. The stack also includes TypeScript, Node.js, npm.
Use freely for any purpose, including commercial use, as long as you keep the copyright notice.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Verify against the repo before relying on details.