A Claude Code plugin providing BrowserAct skills for browser automation and reusable website-skill generation via the browser-act CLI.
Homepage: https://www.browseract.com
| Skill | Purpose |
|---|---|
browser-act |
Runs a browser to extract rendered website data, interact with pages, and capture screenshots. |
browser-act-skill-forge |
Creates reusable website automation Skills with browser-act after authorized exploration. |
browser-act uses the BrowserAct CLI to extract rendered website data, perform page interactions, and capture screenshots. It loads the matching runtime workflow with:
browser-act get-skills core --skill-version 2.0.2browser-act-skill-forge explores a website's implementation path and generates reusable, site-specific Skill packages. Its references/ directory is part of the Skill and defines the exploration procedures and output format.
- Python 3.12
uvbrowser-act-cli
Install the CLI when it is not already available:
uv tool install browser-act-cli --python 3.12Add the marketplace:
/plugin marketplace add browser-act/claude-code-browser-act
Then install the plugin:
/plugin install browser-act@browseract
/plugin marketplace update browseract
The current plugin version is 0.1.0. Versioning follows semver via the version field in plugins/browser-act/.claude-plugin/plugin.json.
.claude-plugin/marketplace.jsondefines the Claude Code marketplace.plugins/browser-act/.claude-plugin/plugin.jsondefines the plugin metadata.plugins/browser-act/skills/browser-act/contains the browser automation Skill.plugins/browser-act/skills/browser-act-skill-forge/contains the reusable website-Skill generation workflow and its required references.README.mdandLICENSEdocument the package and license.
MIT — see LICENSE.