browser-automation · different-ai — Claude Code skill: reliable browser automation
A skill for composable browser automation using minimal OpenCode Browser primitives. Enables agents to click elements, fill forms, and verify state through indexed selector queries with automatic waits.
MIT 1 installs ★ 256 updated 2026-04-03
npx skills add different-ai/agent-bank --skill browser-automation -a claude-codeWhat browser-automation does
The skill provides a safe workflow for browser tasks. Rather than brittle CSS selectors, it uses browser_query with index-based selection—find all matching elements via mode=list, then interact with the target by number. State changes are confirmed after each action using browser_query or browser_snapshot.
Tools include browser_navigate for link traversal, browser_select for native dropdown elements, and browser_open_tab to create tabs. Query modes cover text (visible text), value (input values), list (multiple matches), exists (presence check), and page_text (full page text).
When a selector doesn't match, run browser_query with mode=page_text to verify the content is present. For broad selectors like button or *[role="button"], prefer mode=list and select by index.
What it helps with
- click an element and verify the result changed
- fill a form by selecting an option from a dropdown
- wait for an element to appear with a timeout, then read its text
- navigate between multiple browser tabs
- extract visible page text to check application state
Example prompts for the agent
- Find all buttons on the page and click the third one
- Fill an input field with text and select a value from a dropdown list
- Wait 2 seconds for an element with class
loaderto appear, then get its text
How to install browser-automation
1Skills CLI (Vercel)
npx skills add different-ai/agent-bank --skill browser-automation -a claude-code2Manually: put the files in place
~/.claude/skills/browser-automation/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/browser-automation/
1Skills CLI (Vercel)
npx skills add different-ai/agent-bank --skill browser-automation -a codex2Manually: put the files in place
.agents/skills/browser-automation/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add different-ai/agent-bank --skill browser-automation -a cursor2Manually: put the files in place
.cursor/skills/browser-automation/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install different-ai/agent-bank .opencode/skill/browser-automation2Manually: put the files in place
.github/skills/browser-automation/SKILL.md1Skills CLI (Vercel)
npx skills add different-ai/agent-bank --skill browser-automation -a gemini-cli2Platform CLI
gemini skills install https://github.com/different-ai/agent-bank --path .opencode/skill/browser-automationManually: put the files in place
.devin/skills/browser-automation/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add different-ai/agent-bank --skill browser-automation -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about browser-automation
How do I install browser-automation in Claude Code?
Run npx skills add different-ai/agent-bank --skill browser-automation -a claude-code, then check ~/.claude/skills/browser-automation/SKILL.md to verify.
What does browser-automation do?
It automates browser interactions: clicking, form filling, navigation, and state verification. It uses simple index-based selectors instead of fragile CSS, waits automatically for elements, and confirms results after each action.
What if a normal selector doesn't work?
Use browser_query with mode=list on a broad selector (like button or *[role="button"]) to list all matches with indices, then interact with the target number.
Source text
SKILL.md · MIT
---
name: browser-automation
description: Reliable, composable browser automation using minimal OpenCode Browser primitives.
license: MIT
compatibility: opencode
metadata:
audience: agents
domain: browser
---
## What I do
- Provide a safe, composable workflow for browsing tasks
- Use `browser_query` list and index selection to click reliably
- Confirm state changes after each action
## Best-practice workflow
1. Inspect tabs with `browser_get_tabs`
2. Open new tabs with `browser_open_tab` when needed
3. Navigate with `browser_navigate` if needed
4. Wait for UI using `browser_query` with `timeoutMs`
5. Discover candidates using `browser_query` with `mode=list`
6. Click, type, or select using `index`
7. Confirm using `browser_query` or `browser_snapshot`
## Selecting options
- Use `browser_select` for native `<select>` elements
- Prefer `value` or `label`; use `optionIndex` when needed
- Example: `browser_select({ selector: "select", value: "plugin" })`
## Query modes
- `text`: read visible text from a matched element
- `value`: read input values
- `list`: list many matches with text/metadata
- `exists`: check presence and count
- `page_text`: extract visible page text
## Opening tabs
- Use `browser_open_tab` to create a new tab, optionally with `url` and `active`
- Example: `browser_open_tab({ url: "https://example.com", active: false })`
## Troubleshooting
- If a selector fails, run `browser_query` with `mode=page_text` to confirm the content exists
- Use `mode=list` on broad selectors (`button`, `a`, `*[role="button"]`) and choose by index
- Confirm results after each action
© different-ai · MIT · different-ai/agent-bank
Similar skills
All skills for Agent setup: memory, context, orchestration →browser-automation by other authors comes first.