OIP system: Browser automation
Browser automation
A generated article for one OIP shelf. It lists every operation in this API/CLI/MCP/device/model/core subsystem, links each leaf article, and gives the ledger path for proof.
This page is the operating article for one build subsystem. It is generated from live directory rows. If a task belongs to this subsystem, scan the operations below, open the matching capability article, run only the exact object named there, and verify by receipt.
Kind: device. Capabilities: 9. Machine system map: /api/dispatch?map=BROWSER&format=markdown. Root: /a/oip.
Operations
BROWSER_PLAYWRIGHT
Run a Playwright CLI command on the Mac (npx -y playwright …). g. "screenshot https://x.com /tmp/x.png". First run downloads browsers (~2 min). For full browser automation absorb the playwright MCP server via MCP_PROBE Use when: you need to browser playwright Arguments: the playwright arguments, e. Human article: /a/oip-capability-browser-playwright. Machine doc: ?key=BROWSER_PLAYWRIGHT&format=markdown. Invocation history: /api/invocations?object_id=BROWSER_PLAYWRIGHT.
BROWSER_USE
Run the browser-use agent (AI browser driver). NOT INSTALLED YET — pip3 install --user browser-use + an API key in the Mac env makes this row live Use when: you need to browser use Arguments: task. Human article: /a/oip-capability-browser-use. Machine doc: ?key=BROWSER_USE&format=markdown. Invocation history: /api/invocations?object_id=BROWSER_USE.
BROWSER_MARKDOWN
Get the markdown of a URL via Cloudflare Browser Rendering. $1=account_id, $2=JSON body {url}. Returns the rendered markdown Use when: "fetch as markdown <url>" or "what does <url> say" Arguments: see content. Human article: /a/oip-capability-browser-markdown. Machine doc: ?key=BROWSER_MARKDOWN&format=markdown. Invocation history: /api/invocations?object_id=BROWSER_MARKDOWN.
BROWSER_SCREENSHOT
Get a PNG screenshot of a URL via Cloudflare Browser Rendering. $1=account_id, $2=JSON body {url, screenshotOptions?}. Returns binary PNG Use when: "screenshot <url>" Arguments: see content. Human article: /a/oip-capability-browser-screenshot. Machine doc: ?key=BROWSER_SCREENSHOT&format=markdown. Invocation history: /api/invocations?object_id=BROWSER_SCREENSHOT.
BROWSER_PDF
Render a URL as PDF via Cloudflare Browser Rendering. $1=account_id, $2=JSON body {url}. Returns binary PDF Use when: "save <url> as PDF" Arguments: see content. Human article: /a/oip-capability-browser-pdf. Machine doc: ?key=BROWSER_PDF&format=markdown. Invocation history: /api/invocations?object_id=BROWSER_PDF.
BROWSER_LINKS
Extract all links from a URL via Cloudflare Browser Rendering. $1=account_id, $2=JSON body {url} Use when: "what links does <url> have" Arguments: see content. Human article: /a/oip-capability-browser-links. Machine doc: ?key=BROWSER_LINKS&format=markdown. Invocation history: /api/invocations?object_id=BROWSER_LINKS.
BROWSER_SCRAPE
Extract structured data by selectors via Cloudflare Browser Rendering. $1=account_id, $2=JSON body {url, elements:{selector}]} Use when: "scrape <selector> from <url>" Arguments: see content. Human article: [/a/oip-capability-browser-scrape. Machine doc: ?key=BROWSER_SCRAPE&format=markdown. Invocation history: /api/invocations?object_id=BROWSER_SCRAPE.
BROWSER_JSON
Extract LLM-structured JSON from a URL via Cloudflare Browser Rendering. $1=account_id, $2=JSON body {url, prompt?, response_format?} Use when: "pull <fields> as json from <url>" Arguments: see content. Human article: /a/oip-capability-browser-json. Machine doc: ?key=BROWSER_JSON&format=markdown. Invocation history: /api/invocations?object_id=BROWSER_JSON.
BROWSER_FETCH
Invokable OIP capability. Human article: /a/oip-capability-browser-fetch. Machine doc: ?key=BROWSER_FETCH&format=markdown. Invocation history: /api/invocations?object_id=BROWSER_FETCH.