Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/stigmavlc/claude-bridge/browser-bridgenpx skills add Stigmavlc/claude-bridge --skill browser-bridgegit clone --depth 1 https://github.com/Stigmavlc/claude-bridgeWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/stigmavlc/claude-bridge/browser-bridge)<a href="https://agentmods.dev/skills/stigmavlc/claude-bridge/browser-bridge"><img src="https://agentmods.dev/badge/skills/stigmavlc/claude-bridge/browser-bridge.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00159 | $0.02096 |
| Opus 5 | $0.00079 | $0.01048 |
| Sonnet 5 | $0.00032 | $0.00419 |
| Haiku 4.5 | $0.00016 | $0.00210 |
Grade A, and why
browser-bridge scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 164 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Browser Bridge
You are the bridge between Claude Code and the browser. When a task has a step that lives online, you do it. You don't tell the user to do it manually and you don't hand them a prompt to paste somewhere else. You drive a real browser, watch what happens, decide the next step, and keep going until the task is done. Then you report what happened in the terminal.
This kills the old loop (ask Claude Code for a prompt → paste into the Claude browser extension → copy its reply → paste back → repeat). Here, Claude Code is the agent and the browser is its hands and eyes, all in one conversation.
Step 0: Should this run in the user's logged-in browser, or a clean one?
Decide this first, per task. The rule:
- LOGIN PATH (use the user's real, signed-in Chrome). The task touches the user's own account or data: opening or changing a dashboard (Google/Meta Ads, hosting/cPanel, DNS, Vercel/Netlify, analytics, Stripe, a CMS, email, social), reading something only visible when signed in, posting/publishing as the user, changing account settings, anything that needs their identity.
- NO-LOGIN PATH (use a clean headless browser). The task is public and anonymous: looking up docs/prices/availability, reading an article, scraping a public page, checking whether a site is up, filling a public form that needs no account.
If you are not sure, default to NO-LOGIN and try it. If you hit a sign-in wall, escalate to the LOGIN PATH (Step 1) rather than guessing credentials.
Never type the user's username, password, OTP, or payment details yourself. Authentication is the user's job; on the login path the real browser is already signed in, so you won't need to.
Step 1: Pick the engine that's actually available
Look at which browser tools exist in this session (by tool-name prefix) and pick in this order for the route you chose. Don't assume; check what's present.
LOGIN PATH (drive the user's real Chrome via the extension):
mcp__claude-in-chrome__*(the Claude in Chrome extension; reuses the user's logins). This is the right tool for account/dashboard work.- If these tools are not present, the integration is off. Tell the user,
once and plainly: "This step needs your logged-in browser. Enable it with
/chrome(or restart Claude Code withclaude --chrome), then tell me to continue." Then stop and wait. Do not fall back to a clean browser for an account task; it won't be signed in.
- If these tools are not present, the integration is off. Tell the user,
once and plainly: "This step needs your logged-in browser. Enable it with
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 6d ago First seen · 164 lines · 159 tokens per session scan A 8af045cfdb5b
browser-bridge is a skill published in the GitHub repository Stigmavlc/claude-bridge (2 stars, last pushed 23d ago), licensed MIT. It adds 159 tokens to every session and 2,096 once invoked, about $0.0008 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
surf
Control Chrome browser via CLI for testing, automation, and debugging. Use when the user needs browser automation, screenshots, form filling, page inspection, network/CPU emulation, DevTools streaming, or AI queries via ChatGPT/Gemini/Perplexity/Grok/AI Studio.
control-chrome-cdp
Start or control one exact existing Chrome profile through the Agent OS Chrome CDP bridge with Codex-compatible browser operations. Use for tasks that depend on logged-in Chrome state, including when no Chrome window is open, background tab control, raw CDP, screenshots, page-exposed media saving, or page File…
codex-control-chrome-mcp
Use Codex Control Chrome MCP to control the user's normal Chrome profile through the Codex Chrome extension. Use when a task needs existing Chrome tabs, cookies, logged-in sessions, installed extensions, CDP commands, page inspection, navigation, screenshots, network/console events, or browser automation from an MCP…
reviewing-flows
Use when the pending flow queue needs triage before approval, such as deciding which recorded flows are worth keeping, clearing out ones that can never replay, or reading a flow's real steps before consenting to it.
core
Core chrome-use usage guide. Read this before running any chrome-use commands. Covers the snapshot-and-ref workflow, navigating pages, interacting with elements (click, fill, type, select), extracting text and data, taking screenshots, managing tabs, handling forms and auth, waiting for content, running multiple…
browser-macros
Use when a browser task resembles a repeated flow or a reusable browser script may already cover the requested result.