What is tidesurf-browser?

A tool for controlling a real Chromium web browser from an agent. It shows pages as compact text with IDs for links, buttons, and form fields, which the agent can then click, fill in, select, or scroll.

Use it to browse websites, fill forms, click controls, select options, scroll pages, manage tabs, and take screenshots.

It lets an agent work with live websites, including pages that need interaction, while limiting how much page content must be read at once.

tidesurf-browser — Skill for Claude Code and Codex by TideSurf/core

Install

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.

Any agent

npx skills add TideSurf/core --skill tidesurf-browser

Clone the repo

git clone --depth 1 https://github.com/TideSurf/core

Made for: Claude Code, Codex.

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/skills/tidesurf/core/tidesurf-browser/github.svg)](https://agentmods.dev/skills/tidesurf/core/tidesurf-browser)

Your own site

<a href="https://agentmods.dev/skills/tidesurf/core/tidesurf-browser"><img src="https://agentmods.dev/badge/skills/tidesurf/core/tidesurf-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.

Your own site · 80×15

<a href="https://agentmods.dev/skills/tidesurf/core/tidesurf-browser"><img src="https://agentmods.dev/badge/skills/tidesurf/core/tidesurf-browser.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 97 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 1,235 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits

NVIDIA SkillSpector pass 7 Sept 2026

How audits are shown

Origin original No closer match found in the catalogue.

Token cost

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	$0.00097	$0.01235
Opus 5.5	$0.00039	$0.00494
Sonnet 5.5	$0.00019	$0.00247
Haiku 4.5	$0.00010	$0.00123

Measured 2mo ago against content hash 1c51a070d5b4, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

tidesurf-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 2mo 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.

skills/tidesurf-browser/SKILL.md · 86 lines

How it starts

The opening of the file, as written. The whole thing — 86 lines — stays where its author put it; the contents beside it link to each section on GitHub.

TideSurf Browser

TideSurf gives you a real Chromium browser. It launches a browser or attaches to a running one over CDP, then compresses the live DOM into token-efficient text annotated with action IDs. You read the page as text, act on elements by ID, and re-read to see the result.

The full input table for every tool is in references/tool-reference.md.

The core loop

Every browser task follows the same loop:

get_state to read the current page as compact text. Interactive elements are tagged with IDs like L1 (link), B2 (button), I3 (input).
Pick the element ID that matches your intent.
Act on it: click, type, select, or scroll.
Re-read with get_state to observe the result.

Element IDs are snapshot-scoped. An ID is valid only for the snapshot that produced it. Navigation, DOM mutations, and most actions invalidate prior IDs. Never reuse an ID from an earlier snapshot; call get_state again and use the fresh IDs. This applies even when the page "looks the same" — re-rendered nodes get new IDs.

Token discipline

Page state is the main context cost. Control it:

Pass maxTokens to get_state to cap the snapshot size. Start small and raise it only if the element you need is missing.
Pass viewport: true to restrict the snapshot to what is on screen.
Use mode to shape the output: full for the complete annotated DOM, minimal for structure only, interactive for actionable elements only. Prefer interactive when you only need to act.
On huge pages, run search with a query first to locate the relevant region instead of dumping full state.
Use includeHidden only when the element you need is not visible (e.g. collapsed menus).

Forms

type enters text into an input by ID. Pass clear: true to replace existing content instead of appending.
select sets a <select> element by ID to a given option value.
upload attaches a local file to a file input by ID. filePath must be a path on the local filesystem.
After submitting a form, treat the page as changed: re-read state before doing anything else.

Read the full file on GitHub · 86 lines

Files

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.

references/tool-reference.md 1.8 KB

Changes

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.

2mo ago First seen · 86 lines · 97 tokens per session scan A 1c51a070d5b4

Subscribe to this mod's changes

tidesurf-browser is a skill published in the GitHub repository TideSurf/core (23 stars, last pushed 9d ago), licensed Apache-2.0. It adds 97 tokens to every session and 1,235 once invoked, about $0.0004 per session on Opus 5.5. 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 2026-08-30.

Browser automation ai-agents browser-automation browser-use browser-use-agent cdp

tidesurf-browser next to 6 skills from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
tidesurf-browserTideSurf/core	Skill	23	9d ago	Apache-2.0	A	97	Claude Code, Codex
ChromeiFurySt/open-browser-use	Skill	414	3d ago	MIT	A	34	Claude Code, Codex
open-browser-useiFurySt/open-browser-use	Skill	414	3d ago	MIT	A	97	Codex
browser_harnessbrowser-use/browser-harness	Skill	19k	yesterday	MIT	A	0	Claude Code
browser-use-terminalbrowser-use/terminal	Skill	660	6d ago	MIT	A	41	Claude Code, Codex
browser-harnessbrowser-use/browser-harness	Skill	19k	yesterday	MIT	A	46	Claude Code
x402browser-use/browser-use	Skill	120k	yesterday	MIT	A	175	Claude Code