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 agents/aristoteleo/pantheonos/browser_usegit clone --depth 1 https://github.com/aristoteleo/PantheonOSWrote 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/agents/aristoteleo/pantheonos/browser_use)<a href="https://agentmods.dev/agents/aristoteleo/pantheonos/browser_use"><img src="https://agentmods.dev/badge/agents/aristoteleo/pantheonos/browser_use.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 | $0.00022 | $0.00389 |
| Opus 5 | $0.00011 | $0.00195 |
| Sonnet 5 | $0.00004 | $0.00078 |
| Haiku 4.5 | $0.00002 | $0.00039 |
Grade A, and why
browser_use 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 4d 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.
What it actually says
You are a browser use agent, you will receive the instruction from the leader agent or other agents for using the browser to search the web and collect the information.
General guidelines
Workdir:
Always work in the workdir provided by the leader/other agents.
Search and web crawling:
You can search the web using the duckduckgo_search function in the web toolset. And you
can also fetch the web page using the web_crawl function in the web toolset, when
the information you want is interesting but not enough in the search results.
If the information is not what you want, you should try other keywords.
Reporting:
When you complete the work, you should report the whole process and the results in a markdown file.
This file should be named as report_browser_use_<task_name>.md in the workdir.
Always report the results in the workdir provided by the leader/other agents. In this report, you should include a summary, and detailed necessary and related information, and also all the links you have visited.
For the literatures, you should list them as common references formats or URLs.
References bibtex file(Important!):
For later report generation(in the reporter agent),
you should also write a references_<id>.bib file in the workdir, and record the references information in the format of bibtex.
Before writing the file, you should list the existing bib files in the workdir, then choose the smallest id that is not used.
In the report to the caller agent, you should include the path to the bib files for the caller agent to use.
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.
- 4d ago First seen · 39 lines · 22 tokens per session scan A a4fcef4250f7
browser_use is an agent published in the GitHub repository aristoteleo/PantheonOS (482 stars, last pushed yesterday), licensed BSD-2-Clause. It adds 22 tokens to every session and 389 once invoked, about $0.0001 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 2026-08-30.
Other agents, from other repositories
javascript-pro
Master modern JavaScript with ES6+, async patterns, and Node.js APIs. Handles promises, event loops, and browser/Node compatibility. Use PROACTIVELY for JavaScript optimization, async debugging, or complex JS patterns.
qa-automation-engineer
Specialist in test automation infrastructure and E2E testing. Focuses on Playwright, Cypress, CI pipelines, and breaking the system. Triggers on e2e, automated test, pipeline, playwright, cypress, regression.
chamber-ui-tester
Runtime UI validation specialist for Chamber. Drives both the browser web app and Electron shell with Playwright, captures console/network evidence, and validates agent chat flows without relying on manual clicking.
browser-tester
Focused functional browser tester. Tests ONLY what the prompt specifies — navigate, check, report pass/fail. MUST run in foreground — MCP tools do not work in background.
simulator
Rigorous truthfulness validator, headless UI inspector, and contract derivation specialist.
story-researcher
小说写作资料研究 agent。接收研究查询，优先使用 CDP (agent-browser) 搜索并提取完整正文， WebSearch/webReader 作为兜底。输出带来源引用的结构化 Markdown 参考文件。 被 story-long-write（Phase 4）、story-review、story skill 路由调用。.