Skip to content
Agentmart
Submit a tool

browser-mission

Drive headless Chromium via Playwright MCP for JS sites, logins, and visual checks.

Skill by Raunak-dev-18

IndexedMIT
Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/raunak-dev-18-browser-mission

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Drive headless Chromium via Playwright MCP for JS sites, logins, and visual checks. Use for any web task beyond a simple fetch.

Use cases: testing, productivity, agent-building

Reviews

Write a review

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/browser-mission

What's inside

Skill name
browser-mission
Description
Drive headless Chromium via Playwright MCP for JS sites, logins, and visual checks. Use for any web task beyond a simple fetch.
License
MIT
Instructions
24 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

# Browser missions (Playwright MCP)

You have `browser_*` tools on a headless Chromium with vision.

## Discipline (non-negotiable)

1. `browser_navigate` → `browser_snapshot` FIRST. Never act blind.
2. Prefer accessibility references (`ref`) over coordinates.
3. After navigation/clicks that load content: `browser_wait_for` (load or networkidle, short timeout).
4. Read titles/prices/text from snapshots, not guesses.
5. Screenshot (`browser_screenshot` → save to OUTBOX) to verify ANY visual claim before reporting it.
6. Forms: fill every required field, then click; if login fails, screenshot and report exactly what the page shows.

## Credentials

- Never invent logins. If credentials are missing, ask via :::ASK.
- Never print passwords/tokens in chat; confirm only "logged in OK".

## Limits

- No live desktop exists — browser + shell only.
- Heavy bot-walled sites (some video/streaming walls): try once with waits; if hard-blocked (captcha/consent loop), say BLOCKED + what you see, and offer the closest alternative (official API, RSS, search snippets).

From Raunak-dev-18/tg-bot-oc-machine, MIT.

More from Raunak-dev-18/tg-bot-oc-machine

Signals

Stars
0
Forks
0
Open issues
0
Last commit (this folder)
Oct 12, 2026 (16 days ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
Repository
github.com/Raunak-dev-18/tg-bot-oc-machine
First seen
Oct 24, 2026
Last checked
Oct 24, 2026
Source
GitHubgithub-graphql

More from Raunak-dev-18

Cite this page

browser-mission skill. Agent Mart. https://agentmart.dev/skills/raunak-dev-18-browser-mission (last updated 2026-10-24)