项目文件夹
您已经派生过 browser-use--browser-harness
sauravpanda
0b76d7a8e5
All skills written after running actual browser-harness sessions against live sites. Real selectors, real timings, real gotchas — not theoretical. github/scraping.md: - Confirmed REST API fields, rate limits (60/hr core, 10/min search) - search/code needs auth (401 unauthenticated) - article.Box-row works on trending; needs wait(2) after wait_for_load() for React hydration - raw.githubusercontent.com has no rate limit hackernews/scraping.md: - athing submission (not athing) to exclude comment rows - Job posts have no score row — breaks naive zip; detect by None score - html.unescape() required for titles - Algolia API confirmed; Firebase topstories order matches front page amazon/product-search.md: - new_tab() required on first connection (goto silently fails) - .zg-item-immersion gone from Best Sellers — use [data-asin]+[id="gridItemRoot"] - #priceblock_ourprice is legacy/null — use .a-price-whole + .a-price-fraction - Review count: [aria-label*="ratings"] not .s-underline-text (cross-sell widget collision) news-aggregation/multi-source.md: - The Verge is Atom not RSS — needs namespace dict, not .//item - Reuters hard-blocks http_get (403) even with User-Agent - BBC shows no consent banner from US IP - Parallel fetch 4.3x faster than sequential (0.16s vs 0.70s for 7 feeds) producthunt/scraping.md: - goto() ERR_ABORTED on PH — always use new_tab() - a[href^="/posts/"] matches zero — URLs are /products/ not /posts/ - [data-test^="post-item-"] is the correct card selector - Homepage has 30 fixed items, no lazy load - /leaderboard/daily/YYYY/M/D for complete daily data
Browser Harness
The simplest, thinnest, and most powerful harness to control your real browser with your agent.
Setup prompt
Set up https://github.com/browser-use/browser-harness for me.
Read `install.md` first to install and connect this repo to my real browser. Then read `SKILL.md` for normal usage. Always read `helpers.py` because that is where the functions are. When you open a setup or verification tab, activate it so I can see the active browser tab. After it is installed, if I am already logged in to GitHub, star this repository as a small verification task; if I am not logged in, just go to browser-use.com.
When this page appears tick the checkbox so the agent can connect to the real browser.
Example task
Star this repository.
See domain-skills/ for examples on other websites.
Code Structure
install.mdexplains first-time install and browser bootstrap.SKILL.mdexplains day-to-day browser harness usage.run.py(~13 lines) executes plain Python with helpers preloaded.helpers.py(~192 lines) holds the starting tool calls to interact with the browser and can be modified by the agent.admin.py(~139 lines) holds daemon bootstrap and optional remote-browser helpers.daemon.py(~220 lines) keeps the CDP websocket and socket bridge alive.
Optional: Remote browsers
Useful for sub-agents or deployment.
- Get a Browser Use API key: cloud.browser-use.com/new-api-key
- The agent can also sign up by itself by fetching docs.browser-use.com/llms.txt, which contains the setup flow and challenge context for getting a Browser Use API key.