browser
/browser
Browse the web in a real Chromium: open pages, read them, click, fill forms, take screenshots. Use when a page needs JavaScript, interaction or a screenshot (plain reading of static pages is `web_fetch`).
- Author
- xarantolus ↗
- Repository
- botler ↗
- Licence
- MIT
- Last repository change
- 13.11.2026
Assessment and security
- Files in this package
- 1
- Scripts
- 0
- HTML comments
- 0
- URLs in SKILL.md
- 3
No signals detected. Repository changed on 13.11.2026, without references to retired models or APIs.
MIT, open: use and modify while retaining author attribution.
What it does and when to use it
Browse the web in a real Chromium: open pages, read them, click, fill forms, take screenshots. Use when a page needs JavaScript, interaction or a screenshot (plain reading of static pages is `web_fetch`).
Original author description
Browse the web in a real Chromium: open pages, read them, click, fill forms, take screenshots. Use when a page needs JavaScript, interaction or a screenshot (plain reading of static pages is `web_fetch`).
How it works
- Getting it
- Using it
- What to expect
- From a job (deterministic scraping)
SKILL.md preview
---
name: browser
description: Browse the web in a real Chromium: open pages, read them, click, fill forms, take screenshots. Use when a page needs JavaScript, interaction or a screenshot (plain reading of static pages is `web_fetch`).
---
# The browser
`jobs/browser` is a real desktop Chromium (not headless, so fewer sites block it) with ads and
trackers blocked, driven through [Playwright MCP](https://github.com/microsoft/playwright-mcp).
## Getting it
If you don't have `browser_*` tools, call `request_access(resource="service:browser", reason=...)`.
It is granted without asking the user. You get the tools when you wake, after this turn.
## Using it
- `browser_navigate(url)`, then `browser_snapshot()`: the page as an accessibility tree. Each
element has a `ref`, which you pass to `browser_click`, `browser_type`, `browser_fill_form` or
`browser_select_option`. Prefer snapshots to screenshots for reading: they are text and much
cheaper.
- `browser_take_screenshot()` when the user should *see* the page. The screenshot shows in the
chat. `attach_files` sends a saved file to another chat.
- `browser_tabs` lists, opens and closes tabs; `browser_wait_for` waits for text or a delay;
`browser_evaluate` runs JavaScript in the page; `browser_console_messages` and
`browser_network_requests` help when something does not load.
## What to expect
- **Your session is yours alone.** Your connection has its own browser context, with its own
cookies, storage and tabs. Other chats using the browser at the same time never see them, and
you never see theirs.
- **It does not last.** The context closes after 10 minutes without a browser tool call, and
whenever you sleep. You then start fresh and logged out. Finish a flow in one go, and keep
anything you need (text, a screenshot path) in your home.
- **Logins:** for a site the user stays logged into over time (a portal, webmail), don't type
their password here. Use skill `browser-login`: a separate service keeps that session, and the
user logs in once through a visit.
- **Blocked domains:** ad and tracker domains don't resolve. If a page breaks because of it,
that is the list (`images/browser`, pinned StevenBlack hosts). Say so rather than working
around it.
## From a job (deterministic scraping)
A job or service can use the same browser: `services = ["browser"]` in its `job.toml` (the
first time asks the user), then talk MCP to `http://browser:8000/mcp`. Each MCP session is its
own context, so a job never disturbs chats. In Python, the `mcp` package's client:
```python
from mcp import ClientSession
from mcp.client.streamable_http import streamablehttp_client
async with streamablehttp_client("http://browser:8000/mcp") as (read, write, _):
async with ClientSession(read, write) as s:
await s.initialize()
await s.call_tool("browser_navigate", {"url": url})
page = await s.call_tool("browser_evaluate", {"function": "() => document.body.innerText"})
```
Keep the parsing, dedupe (cursor in `botler.state`) and typed events in the job. The browser only
loads pages.
Versions
| Repository | Score | Licence |
|---|---|---|
| This page · xarantolus/botler | 100 | MIT |
| vibiumdev/vibium | 100 | Apache-2.0 |
| superset-sh/superset | 100 | MIT |
| phnx-labs/agi-cli | 100 | Apache-2.0 |
| phnx-labs/browser-cli | 100 | Apache-2.0 |
| Embervy-ai/embervy | 100 | MIT |
| BlackBeltTechnology/pi-agent-dashboard | 100 | MIT |
| shibing624/agentica | 100 | Apache-2.0 |
| iammrduncan/fxbot | 100 | Apache-2.0 |
Database snapshot: 13.11.2026 · package version d72aa6859b12 (SHA-256 digest of the entire package)
Downloads in the last 30 days: 0 · ranking
Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.
Similar skills
- 92ai-browser
Drive Chrome AI, the user's dedicated signed-in Chrome window, with the dev-browser CLI to open sites, read pages, click, fill forms, take screenshots and check a live app. Use when the user asks to browse, open a site, click, fill a form, take a screenshot, scrape a page, test… - 100web-browse
Drive a real Chromium browser step by step from the command line on the user's own machine - a session that persists between commands, open pages, a snapshot of elements with numbered refs, click, type, select, press keys, scroll, wait, tabs, read the page as Markdown (via… - 100playwright-cli
Automate browser interactions, test web pages and work with Playwright tests using the playwright-cli command. Use when the user wants to open a browser, navigate pages, click elements, fill forms, take screenshots, run Playwright tests, debug browser automation, or perform any… - 90Browser Automation
This skill should be used when the user needs to interact with web pages, browse websites, take screenshots, fill forms, click elements, extract web content, or perform any browser automation tasks. Trigger phrases include "open webpage", "visit website", "browse", "screenshot… - 95agent-browser
Browse the web for any task — research topics, read articles, interact with web apps, fill forms, take screenshots, extract data, and test web pages. Use whenever a browser would be useful, not just when the user explicitly asks. - 100browse
Browse a web page with a fast headless browser: open it, read it, click through a flow, take screenshots, or scrape its data. Use when the person says open this page, test the site, take a screenshot, dogfood this, or scrape this page. Runs on the gstack browse engine, an…