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 skills add joonlab/joonlab-claudecode-setting-for-share --skill cmux-browsergit clone --depth 1 https://github.com/joonlab/joonlab-claudecode-setting-for-shareWrote 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/joonlab/joonlab-claudecode-setting-for-share/cmux-browser)<a href="https://agentmods.dev/skills/joonlab/joonlab-claudecode-setting-for-share/cmux-browser"><img src="https://agentmods.dev/badge/skills/joonlab/joonlab-claudecode-setting-for-share/cmux-browser/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/joonlab/joonlab-claudecode-setting-for-share/cmux-browser"><img src="https://agentmods.dev/badge/skills/joonlab/joonlab-claudecode-setting-for-share/cmux-browser.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00183 | $0.02742 |
| Opus 5 | $0.00092 | $0.01371 |
| Sonnet 5 | $0.00037 | $0.00548 |
| Haiku 4.5 | $0.00018 | $0.00274 |
Grade A, and why
cmux-browser scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 224 lines — stays where its author put it; the contents beside it link to each section on GitHub.
cmux browser
Mandatory Behavior
When this skill is activated:
- MUST use
cmux browsercommands via Bash tool for ALL web interactions - MUST NOT use WebFetch, WebSearch, or any
mcp__claude-in-chrome__*tools as substitutes - If cmux browser fails after 3 attempts, report the error to the user instead of silently switching tools
- Sub-agents CAN and SHOULD use cmux browser — see Parallel Browsing section
Surface Model
Every session starts with open, which returns a surface ID. Use it in all subsequent commands.
cmux browser open https://example.com # → OK surface=surface:7
cmux browser surface:7 <subcommand> [args] [flags]
Core Workflow
open → wait → snapshot/extract → interact → verify
Note: close subcommand is NOT supported. Surfaces are cleaned up automatically.
Task Recipes
Recipe 1: Read & Summarize a Web Page
# 1. Open — macOS-safe surface parse (BSD grep has no -P / \K)
SURFACE=$(cmux browser open https://example.com 2>&1 | grep -oE 'surface:[0-9]+' | head -1)
# 2. Wait for load (prefer --selector/--text when a specific element is needed)
cmux browser $SURFACE wait --load-state complete --timeout-ms 15000
# 3. Extract text — PREFER get text / get html over eval (see Known Limitations)
cmux browser $SURFACE get text "body"
cmux browser $SURFACE get text "main" # or "article"
cmux browser $SURFACE get html "body" # NOTE: get html REQUIRES a selector
For long pages, scroll and re-extract:
cmux browser $SURFACE get text "body"
cmux browser $SURFACE scroll --dy 2000
cmux browser $SURFACE get text "body"
Recipe 2: Extract Structured Data
Prefer DOM-access commands — they work even where eval is blocked (see Known Limitations).
cmux browser $SURFACE get attr "a.primary" --attr href
cmux browser $SURFACE get count ".list-item" # may js_error on CSP sites → fall back to get html + parse
cmux browser $SURFACE get html "body" # then parse with shell/python (grep, jq, BeautifulSoup)
# eval works on MOST sites, but FAILS with js_error on CSP-strict sites (e.g. Hacker News):
cmux browser $SURFACE eval "JSON.stringify(Array.from(document.querySelectorAll('h2')).map(h => h.textContent))"
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 10d ago First seen · 224 lines · 183 tokens per session scan A 2f2b35373bda
cmux-browser is a skill published in the GitHub repository joonlab/joonlab-claudecode-setting-for-share (10 stars, last pushed 1mo ago), licensed MIT. It adds 183 tokens to every session and 2,742 once invoked, about $0.0009 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
opencli-sitemap-author
Use when creating or maintaining OpenCLI site sitemaps: agent-facing navigation, page-state, action, workflow, API-reference, pitfall, and fallback knowledge for a website. Use after browser exploration discovers durable site context, when a sitemap is stale, or when promoting local site knowledge into the repo.
debug-optimize-lcp
Guides debugging and optimizing Largest Contentful Paint (LCP) using Chrome DevTools MCP tools. Use this skill whenever the user asks about LCP performance, slow page loads, Core Web Vitals optimization, or wants to understand why their page's main content takes too long to appear. Also use when the user mentions…
interactive-login
How to complete browser/interactive logins (aws / gh / glab / gcloud). The platform backgrounds the login poller so it survives the human's browser round-trip — and when that does NOT work.
pinchtab-mcp
Use this skill when a task requires browser automation through PinchTab's MCP server connected to a remote browser instance. Covers navigation, element interaction, data extraction, form filling, multi-step flows, and session management via MCP tools.
azure-messaging-webpubsub-java
Build real-time web applications with Azure Web PubSub SDK for Java. Use when implementing WebSocket-based messaging, live updates, chat applications, or server-to-client push notifications.
google-safe-browsing
Prevent and fix Google Safe Browsing "Dangerous site" flags. Use when launching a public web app, buying/picking a domain, building a login or signup page, or when any site shows a red "Dangerous site" / "Deceptive site" warning in Chrome, Brave, Safari, Firefox, or Edge. Triggers on "dangerous site", "deceptive…