agent-browser
Use when the user wants browser automation against an existing Chrome or Chromium tab, especially to reuse a real logged-in session without reauthenticating.
Skill by kwannoel
Chalk Labs
Need help deploying agent-browser?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
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/kwannoel-agent-browser
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.
About
Use cases: agent-building
Reviews
Write a reviewNo 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
- Runs as
- Shell scripts
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/agent-browser
Tools it pre-approves 3
From allowed-tools in SKILL.md: the agent may use these without asking you each time.
Bash(npxagent-browser:*),Bash(agent-browser:*)
What's inside
- Skill name
agent-browser- Description
- Use when the user wants browser automation against an existing Chrome or Chromium tab, especially to reuse a real logged-in session without reauthenticating. Triggers include requests to use the current browser, keep existing cookies or extensions, avoid login churn, or automate a live Chromium tab.
- License
- MIT
- Allowed tools
- Bash(npx, agent-browser:*),, Bash(agent-browser:*)
- Instructions
- 219 lines in SKILL.md · view on GitHub
Before installing
- Ships executable scripts
- Pre-approves some shell commands
Files (11)
- SKILL.md
- authentication.md · references/
- commands.md · references/
- profiling.md · references/
- proxy-support.md · references/
- session-management.md · references/
- snapshot-refs.md · references/
- video-recording.md · references/
- authenticated-session.sh · templates/
- capture-workflow.sh · templates/
- form-automation.sh · templates/
SKILL.md (first lines)
# Browser Automation with agent-browser `agent-browser` is the Chromium skill. Prefer it when the user wants the agent to work against the browser they already have open and logged into. The primary model is OpenClaw-style relay control: 1. Open the target site in Chrome or Chromium. 2. Attach the browser relay extension to the tab you want the agent to control. 3. Run `agent-browser` commands against that attached tab. 4. Detach when you are done. If relay mode is unavailable, fall back to a managed dedicated browser profile. ## Mode Selection ### Relay mode Use relay mode when: - the user is already logged in - the task depends on the user's real browser extensions, cookies, or in-progress state - the user wants to avoid login churn Relay mode means the agent controls a real Chrome or Chromium tab. Treat that tab as sensitive. Use the minimum scope needed and avoid unrelated tabs. ### Managed mode Use managed mode when: - the user wants isolation from their daily browser - the relay is not installed or not attached - the task needs a dedicated automation profile Managed mode launches or reuses a dedicated browser profile instead of the user's live browsing tab. ## Core Workflow Every `agent-browser` task follows this pattern: 1. Confirm the right tab is attached in relay mode, or start managed mode. 2. Snapshot the page: `agent-browser snapshot -i` 3. Interact with refs: `click`, `fill`, `select`, `check` 4. Re-snapshot after navigation or major DOM changes ```bash # Attached tab is already open in Chrome/Chromium agent-browser snapshot -i # Output: @e1 [input type="email"], @e2 [input type="password"], @e3 [button] "Sign in" agent-browser fill @e1 "user@example.com" agent-browser fill @e2 "password123" agent-browser click @e3 agent-browser wait --load networkidle agent-browser snapshot -i ``` Do not close the browser in relay mode unless the user explicitly asks for that. In relay mode, prefer detach over `agent-browser close`. ## Quick Start ### Relay mode quick start ```bash # 1. User opens the target site in Chrome/Chromium # 2. User attaches the relay extension on the target tab # 3. Inspect the attached tab agent-browser snapshot -i # 4. Interact agent-browser click @e1 agent-browser wait --load networkidle agent-browser screenshot current-tab.png ``` ### Managed mode quick start ```bash agent-browser --profile ~/.agent-browser/profiles/work open https://example.com agent-browser wait --load networkidle agent-browser snapshot -i ``` ## Authentication When a site requires login, prefer these approaches in this order: ### Option 1: Existing logged-in browser tab This is the default. If the user already has the site open and authenticated in Chrome or Chromium, attach relay control to that tab and continue. This keeps the real session, cookies, local storage, browser extensions, and any in-progress state. ### Option 2: Managed dedicated profile ```bash # First run: log in once agent-browser --profile ~/.profiles/myapp open https://app.example.com/login # Future runs: reuse the same dedicated profile agent-browser --profile ~/.profiles/myapp open https://app.example.com/dashboard ``` Use this when relay mode is unavailable or the user wants isolation. ### Option 3: Saved or imported state Use saved state only when relay mode and dedicated profiles are not a good fit. ```bash # Import auth from an existing compatible browser session agent-browser --auto-connect state save ./auth.json # Reuse saved state agent-browser --state ./auth.json open https://app.example.com/dashboard ``` State files contain session tokens in plaintext unless you add encryption. Keep them out of git and delete them when they are no longer needed. See [references/authentication.md](references/authentication.md) for relay setup, fallback auth patterns, 2FA, and recovery steps.
From kwannoel/the-controller, MIT.
More from kwannoel/the-controller
- 0.485score
the-controller-writing
kwannoel
Remove AI writing patterns from prose.
SkillDFreshness: D★ 13Claude Code +9 - 0.481score
the-controller-e2e-eval
kwannoel
Use when validating UI changes end-to-end before claiming work is complete — spawns a fresh server pair from the worktree and runs Playwright tests against it
SkillDFreshness: D★ 13Claude Code +9 - 0.481score
the-controller-general-e2e-eval
kwannoel
Use when validating UI changes end-to-end in any project before claiming work is complete — runs Playwright tests against the project's dev server to prove…
SkillDFreshness: D★ 13Claude Code +9 - 0.481score
the-controller-validating-agent-applications
kwannoel
Use when building or modifying applications that use AI agents, before shipping or claiming the application works - requires end-to-end validation that…
SkillDFreshness: D★ 13Claude Code +9 - 0.481score
the-controller-finishing-a-development-branch
kwannoel
Use when implementation is complete and you need to merge the branch — verifies tests, rebases, creates PR, squash merges, deletes remote branch, syncs local…
SkillDFreshness: D★ 13Claude Code +9 - 0.481score
the-controller-setting-credentials
kwannoel
Use when an agent needs to set an API key, secret, or credential in a project .env file — ensures secrets never leak into terminal output or conversation…
SkillDFreshness: D★ 13Claude Code +9 - 0.480score
the-controller-autonomous-execution
kwannoel
Use when executing an implementation plan fully autonomously without human interaction
SkillDFreshness: D★ 13Claude Code +9 - 0.477score
the-controller-debugging-ui-with-playwright
kwannoel
Use when a UI behavior doesn't work as expected (key not responding, focus lost, mode not changing) and code reading alone isn't enough — runs Playwright in…
SkillDFreshness: D★ 13Claude Code +9 - 0.476score
the-controller-synchronizing-skills
kwannoel
Use when a new skill has been added to the repo and needs to be symlinked into ~/.claude/skills/ and ~/.codex/skills/custom/ so Claude Code and Codex can…
SkillDFreshness: D★ 13Claude Code +9 - 0.472score
firefox-browser
kwannoel
Firefox browser automation for AI agents using Playwright.
SkillDFreshness: D★ 13Claude Code +9 - 0.472score
the-controller-feedback-widget
kwannoel
Set up an in-app feedback widget with screenshot capture, canvas annotation, and GitHub issue creation.
SkillDFreshness: D★ 13Claude Code +9 - 0.456score
the-controller-receiving-code-review
kwannoel
Use when receiving code review feedback, before implementing suggestions, especially if feedback seems unclear or technically questionable - requires…
SkillDFreshness: D★ 13Claude Code +9
Signals
- Stars
- 13
- Forks
- 2
- Open issues
- 60
- Last commit (this folder)
- Mar 24, 2026 (7 months ago)
- Last release
- Mar 8, 2026 (7 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/kwannoel/the-controller
- Website
- noelkwan.xyz/ctrl
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from kwannoel
- 0.485score
the-controller-writing
kwannoel
Remove AI writing patterns from prose.
SkillDFreshness: D★ 13Claude Code +9 - 0.481score
the-controller-general-e2e-eval
kwannoel
Use when validating UI changes end-to-end in any project before claiming work is complete — runs Playwright tests against the project's dev server to prove…
SkillDFreshness: D★ 13Claude Code +9 - 0.481score
the-controller-setting-credentials
kwannoel
Use when an agent needs to set an API key, secret, or credential in a project .env file — ensures secrets never leak into terminal output or conversation…
SkillDFreshness: D★ 13Claude Code +9 - 0.481score
the-controller-e2e-eval
kwannoel
Use when validating UI changes end-to-end before claiming work is complete — spawns a fresh server pair from the worktree and runs Playwright tests against it
SkillDFreshness: D★ 13Claude Code +9 - 0.481score
the-controller-finishing-a-development-branch
kwannoel
Use when implementation is complete and you need to merge the branch — verifies tests, rebases, creates PR, squash merges, deletes remote branch, syncs local…
SkillDFreshness: D★ 13Claude Code +9 - 0.481score
the-controller-validating-agent-applications
kwannoel
Use when building or modifying applications that use AI agents, before shipping or claiming the application works - requires end-to-end validation that…
SkillDFreshness: D★ 13Claude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get agent-browser running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
agent-browser skill. Agent Mart. https://agentmart.dev/skills/kwannoel-agent-browser (last updated 2026-10-24)