Ultimate Browsing Agent Skill
Escalation skill for blocked or hard-to-reach web access — load it when a normal browse/fetch is blocked (WAF, 403, Cloudflare, JS-only…
- Context cost
- 100k tokensestimated from the bundle, loaded when it triggers
- Bundle
- 56 files · 534.5 kB25 scripts among them — read before you run
- Licence
- not declarednone named by its author
- Last change
- —no release on file
- Servers it uses
- Noneruns standalone
What it does
Escalation skill for blocked or hard-to-reach web access — load it when a normal browse/fetch is blocked (WAF, 403, Cloudflare, JS-only render, login-gated, or a platform a generic fetcher cannot read). Tiered router: TIER 1 insane-search (headless extraction + WAF bypass via curl_cffi TLS impersonation, yt-dlp, Jina Reader, public APIs, Playwright real-Chrome fallback); TIER 1.5 agent-reach (platform-native readers for Chinese and social platforms: Xiaohongshu, Douyin, Weibo, Bilibili, V2EX, WeChat, plus Twitter/Reddit/LinkedIn/GitHub); TIER 2 Chrome stealth (CloakBrowser stealth Chromium +…
Installed, it changes the agent in these ways.
What this skill changes about the agent has not been set out here as a list. Its SKILL.md, shown below, says it in the author’s own words.
And these are the jobs people reach for it for.
The skill itself
A skill is instructions the model reads, so you can see what you are getting: below is the start of its SKILL.md. The front matter loads with every session, the body loads when the skill triggers, and the whole file comes with the install.
Nothing in a skill executes. The client loads the text and the model follows it, so a skill can be audited the way a runbook is — by reading it.
What it costs in context
Skills are not billed by the call. They are paid for in context: every token the instructions occupy is a token your code, your diff and your conversation cannot use. Here is what this one takes and when it takes it.
100k tokens, estimated from the bundle at four bytes to the token, held for the rest of the session once it triggers. Heavy. Teams tend to install this one per project rather than globally, and load it only when the job comes up.
A server charges by the month. A skill charges once per session, in context, and then keeps charging it for as long as the session lives.
Before and after
The same question, put to the same model twice: once as it comes, and once with these instructions loaded.
No worked example has been published for this skill yet.
The procedure it runs
The procedure has not been set out here step by step. It is in the body of its SKILL.md: the start of it is shown above, and the whole file comes with the install.
These steps are written for a model to follow, not executed by a runtime. It can still be told to skip one, and it will say so when it does.
Servers it uses
None. This skill calls no MCP servers at all.
Everything it needs is in the instructions, so it works in a project with nothing connected — the model reads the file and changes how it works with what it can already reach.
Its SKILL.md does not declare which tools it asks the agent for, so what it touches is whatever the instructions themselves ask for — read them before you install it.
This SKILL.md carries no allowed-tools line, so neither is declared. A skill grants no permissions of its own — it can only ask for tools your client already has.
What it will not do
Every skill is narrow, and the useful ones say where they stop. These are the jobs this one is the wrong tool for.
What this skill is not for has not been published here. Nothing is implied by that: it is a section the author has not filled in.
What is in the bundle
56 files, 534.5 kB on disk. Mostly text — the instructions the model reads — with 25 scripts in it that your client would run only if the instructions tell it to. 1 file of its source is not in the download — listed after these, with why.
- SKILL.md10.8 kB
- ATTRIBUTION.md5.0 kB
- engine/AGENTS.md4.1 kB
- engine/__init__.py0.7 kB
- engine/__main__.py4.9 kB
- engine/bias_check.py7.3 kB
- engine/curl_probe.py2.4 kB
- engine/executor.py6.4 kB
- engine/fetch_chain.py13.4 kB
- engine/referers.py0.3 kB
- engine/result_schema.py1.6 kB
- engine/summary.py1.2 kB
- engine/surrogate.py8.1 kB
- engine/surrogates.yaml2.4 kB
- engine/url_transforms.py3.3 kB
- engine/validators.py9.3 kB
- engine/waf_detector.py7.1 kB
- engine/waf_profiles.yaml6.4 kB
- references/chrome-stealth.md6.9 kB
- scripts/cookie_crypto.py3.0 kB
- scripts/cookie_domains.py1.2 kB
- scripts/cookie_paths.py3.8 kB
- scripts/extract_cookies.py9.4 kB
- engine/templates/package.json0.3 kB
- engine/templates/playwright_mobile_chrome.js3.5 kB
- engine/templates/playwright_real_chrome.js5.5 kB
- engine/tests/test_fetch_chain.py2.8 kB
- engine/tests/test_playwright_templates.py11.1 kB
- engine/tests/test_surrogate.py11.6 kB
- engine/tests/test_surrogate_validators.py3.6 kB
- references/agent-reach/README.md2.3 kB
- references/agent-reach/career.md0.7 kB
- references/agent-reach/dev.md1.4 kB
- references/agent-reach/search.md0.8 kB
- references/agent-reach/social.md5.7 kB
- references/agent-reach/video.md2.7 kB
- references/agent-reach/web.md1.9 kB
- references/insane-search/README.md19.0 kB
- references/insane-search/cache-archive.md4.2 kB
- references/insane-search/fallback.md6.2 kB
- references/insane-search/jina.md3.9 kB
- references/insane-search/json-api.md3.6 kB
- references/insane-search/media.md3.6 kB
- references/insane-search/metadata.md3.0 kB
- references/insane-search/naver.md3.4 kB
- references/insane-search/playwright.md6.2 kB
- references/insane-search/public-api.md3.5 kB
- references/insane-search/rss.md3.0 kB
- references/insane-search/tls-impersonate.md6.4 kB
- references/insane-search/twitter.md3.6 kB
- scripts/tests/test_cookie_domain_filter.py3.6 kB
- scripts/tests/test_extract_cookies.py9.6 kB
- engine/tests/fixtures/amp_redirect_stub.html0.3 kB
- engine/tests/fixtures/search_interstitial.html92.5 kB
- engine/tests/fixtures/wayback_available.json0.2 kB
- engine/tests/fixtures/wayback_snapshot.html185.8 kB
Not in the download
Not carried is a type a skill download does not hold. Take it from its source.
- .gitignorenot carried
A skill installs nothing and depends on nothing: it is a folder your client reads. This one carries 25 scripts beside the text, so the download is 56 files — not the whole of its source: 1 file is left out. It is listed under the files, with why.
Install
Installing copies the bundle into your project. Nothing runs at install time — the files sit on disk until the model reads them.
# Ultimate Browsing · 100k tokens when loaded · macOS, Linux
mkdir -p .claude/skills && t=$(mktemp -d) &&
curl -fsSL -o "$t/bundle.tar.gz" https://mcprush.com/api/skills/ultimate-browsing/bundle.tar.gz &&
tar -xzf "$t/bundle.tar.gz" -C "$t" && [ -d "$t/ultimate-browsing" ] &&
{ [ ! -e .claude/skills/ultimate-browsing ] || { mv .claude/skills/ultimate-browsing "$t/.previous" && echo "Previous copy moved to $t/.previous"; }; } &&
mv "$t/ultimate-browsing" .claude/skills/ && echo "Installed .claude/skills/ultimate-browsing"npx mcprush@latest skill add karanmrn/ultimate-browsingWindows PowerShell says 'running scripts is disabled on this system'? Run Set-ExecutionPolicy -Scope CurrentUser RemoteSigned once, or type npx.cmd instead of npx (codex.cmd, gemini.cmd, grok.cmd likewise). PowerShell 7 needs neither. Writes to .claude/skills/ultimate-browsing/ in the current project. For every project instead, run mkdir -p ~/.claude/skills && t=$(mktemp -d) && curl -fsSL -o "$t/bundle.tar.gz" https://mcprush.com/api/skills/ultimate-browsing/bundle.tar.gz && tar -xzf "$t/bundle.tar.gz" -C "$t" && [ -d "$t/ultimate-browsing" ] && { [ ! -e ~/.claude/skills/ultimate-browsing ] || { mv ~/.claude/skills/ultimate-browsing "$t/.previous" && echo "Previous copy moved to $t/.previous"; }; } && mv "$t/ultimate-browsing" ~/.claude/skills/ && echo "Installed ~/.claude/skills/ultimate-browsing" (macOS, Linux) — or add --global to the npx line. The curl line replaces .claude/skills/ultimate-browsing/ only after the download succeeds, so files left there by another skill of the same name, or by an older release, do not mix in. A copy already there is moved aside, not deleted, and the line prints where. Not in this download: 1 file of its source (which one, and why). Take it from its source.
Which clients pick it up on their own
A skill is a folder of text. Most of these read a skills folder and pick it up without being told; the rest take the same folder another way — an upload in the app, a parameter on an API call — and then choose it by its description just the same.
Versions
Its author publishes no version number, so there is nothing here to pin to: what you install is the folder as it stands today. Instructions change more often than APIs do — a skill can be rewritten entirely without anything it depends on moving.
Nothing to pin to: this skill carries no version number of its own. What you install is what the folder holds on the day you install it.
Reviews
Nobody has reviewed this skill yet. The rating is the mean of the reviews written here, so there is none until somebody writes the first.
A review is tied to an account so it can be answered, edited and taken down by the person who wrote it.
Sign inAnybody with an account, one review each. It carries the version that was current when you wrote it, and the publisher may reply once.