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 aksika/abtars --skill browsergit clone --depth 1 https://github.com/aksika/abtarsWrote 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/aksika/abtars/browser)<a href="https://agentmods.dev/skills/aksika/abtars/browser"><img src="https://agentmods.dev/badge/skills/aksika/abtars/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/aksika/abtars/browser"><img src="https://agentmods.dev/badge/skills/aksika/abtars/browser.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00015 | $0.00727 |
| Opus 5.5 | $0.00006 | $0.00291 |
| Sonnet 5.5 | $0.00003 | $0.00145 |
| Haiku 4.5 | $0.00002 | $0.00073 |
Grade A, and why
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 1mo 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 — 77 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Browser Skill
Cloak — the action-capable CLI backed by CloakBrowser's stealth Chromium runtime. It supports persistent sessions, navigation, snapshots, interaction, text extraction, screenshots, cookies, and multi-step workflows.
Role routing
Read your current session type before choosing a path:
- If you are Browsie (session type B / W): continue to Cloak actions below.
- If the user explicitly requests direct Main browser operation in the current turn: follow Emergency Direct Mode.
- Otherwise (Main / any other session): create a B-type kanban card with a detailed goal using
kanban_manage(Direct API) orabtars kanban create(ACP), report the card ID, and stop. Do NOT run browser commands inline.
Browsie — Cloak actions
The first command starts Cloak's local daemon automatically. Create one named session per workflow and reuse it across calls:
cloak session new --name=browse --humanize
Actions
goto— navigate:cloak goto @browse "https://example.com"snapshot— inspect the current page and interactive element UIDs:cloak snapshot @browseclick— click a snapshot UID or selector:cloak click @browse "u7"fill— fill a field:cloak fill @browse "#email" "$EMAIL"text— extract visible text, optionally scoped:cloak text @browse --selector="main"screenshot— capture a page:cloak screenshot @browse --full-page --path="/tmp/page.png"cookies set— load cookies:cloak cookies set @browse --file "/path/to/cookies.json"session close— close the browser session:cloak session close @browse
Use the snapshot UIDs returned by snapshot for reliable interaction. Do not
reuse a UID after navigation; take a fresh snapshot first.
Emergency Direct Mode (Main only)
Use ONLY when the user explicitly states in the current turn that Main should operate the browser directly.
session="main-emergency-$(date +%s)"
cloak session new --name="$session" --humanize
cloak goto "@$session" "https://example.com"
cloak snapshot "@$session"
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.
- 1mo ago Changed · +9 lines 75d613395020
- 1mo ago First seen · 68 lines · 15 tokens per session scan A 3a9696c993ab
browser is a skill published in the GitHub repository aksika/abtars (10 stars, last pushed today), licensed Apache-2.0. It adds 15 tokens to every session and 727 once invoked, about $0.0001 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 2026-08-31.
Other skills, from other repositories
Web Read
Keeps direct URL reading and text extraction available when a specific page matters more than search results.
Web Search
Keeps lightweight public-web discovery available for prompts, scheduled jobs, and follow-up research.
browser
Automate browser interactions for web scraping, testing, and navigation. Use when the user needs to visit a webpage, interact with UI elements, take screenshots, or extract content from rendered pages.
search-craft
How to get real coverage out of a search engine that returns a handful of results per query, and when to reach for the crawler or the browser instead. Use at the start of any search that has to be thorough rather than quick.
firecrawl
Web scraping and crawling for AI agents via Firecrawl MCP. Scrape URLs to markdown, crawl websites, search the web, and extract structured data. Supports cloud API and self-hosted deployments including CRW (Rust alternative).
web-browser
A personal web browser for AI agents.