Skip to content

billy-enrizky/openbrowser-ai / plugin/skills/page-analysis

page-analysis

Flagged

Analyze web page content, structure, and layout to understand what a page contains and how it is organized. Trigger when the user asks to: analyze a page, understand page structure, inspect a website, summarize page content, examine page layout, review a web page, or describe what is on a page.

Page Analysis

Analyze and understand web page content, structure, and interactive elements using Python code execution. Produces a comprehensive breakdown of what is on the page and how it is organized.

All code runs via openbrowser-ai -c. The daemon starts automatically and persists variables across calls. All browser functions are async -- use await.

The CLI daemon also persists cookies and login state in ~/.config/openbrowser/profiles/daemon/storage_state.json, so authenticated sessions can be reused across later runs.

Setup

Before running, verify openbrowser-ai is installed:

openbrowser-ai --help

If not found, install:

# macOS/Linux
curl -fsSL https://raw.githubusercontent.com/billy-enrizky/openbrowser-ai/main/install.sh | sh

# Windows (PowerShell)
irm https://raw.githubusercontent.com/billy-enrizky/openbrowser-ai/main/install.ps1 | iex

Workflow

Step 1 -- Navigate and get overview

openbrowser-ai -c - <<'EOF'
await navigate("https://example.com")
state = await browser.get_browser_state_summary()
print(f"Title: {state.title}")
print(f"URL: {state.url}")
print(f"Interactive elements: {len(state.dom_state.selector_map)}")
print(f"Tabs: {len(state.tabs)}")
EOF

Step 2 -- Extract page metadata

openbrowser-ai -c - <<'EOF'
meta = await evaluate("""
(function(){
  return {
    title: document.title,
    description: document.querySelector("meta[name='description']")?.content,
    canonical: document.querySelector("link[rel='canonical']")?.href,
    ogTitle: document.querySelector("meta[property='og:title']")?.content,
    ogImage: document.querySelector("meta[property='og:image']")?.content,
    lang: document.documentElement.lang,
    charset: document.characterSet
  };
})()
""")

import json
print(json.dumps(meta, indent=2))
EOF

Step 3 -- Detect frameworks and technologies

openbrowser-ai -c - <<'EOF'
tech = await evaluate("""
(function(){
  const t = [];
  if (window.__NEXT_DATA__) t.push("Next.js");
  if (window.__NUXT__) t.push("Nuxt.js");
  if (document.querySelector("[data-reactroot]") || document.querySelector("#__next")) t.push("React");
  if (document.querySelector("[ng-version]")) t.push("Angular");
  if (window.jQuery) t.push("jQuery");
  if (window.Vue) t.push("Vue.js");
  if (document.querySelector("[data-svelte]")) t.push("Svelte");
  return t;
})()
""")
print(f"Technologies detected: {tech}")
EOF

Step 4 -- Content summary and statistics

openbrowser-ai -c - <<'EOF'
stats = await evaluate("""
(function(){
  return {
    headings: document.querySelectorAll("h1,h2,h3,h4,h5,h6").length,
    paragraphs: document.querySelectorAll("p").length,
    images: document.querySelectorAll("img").length,
    links: document.querySelectorAll("a").length,
    forms: document.querySelectorAll("form").length,
    tables: document.querySelectorAll("table").length,
    lists: document.querySelectorAll("ul,ol").length,
    buttons: document.querySelectorAll("button,[role='button']").length,
    inputs: document.querySelectorAll("input,textarea,select").length,
    iframes: document.querySelectorAll("iframe").length,
    scripts: document.querySelectorAll("script").length,
    stylesheets: document.querySelectorAll("link[rel='stylesheet']").length
  };
})()
""")

import json
print("Content statistics:")
print(json.dumps(stats, indent=2))
EOF

Step 5 -- Analyze heading structure

openbrowser-ai -c - <<'EOF'
headings = await evaluate("""
(function(){
  return Array.from(document.querySelectorAll("h1,h2,h3,h4,h5,h6")).map(h => ({
    tag: h.tagName,
    text: h.textContent.trim().substring(0, 80)
  }));
})()
""")

for h in headings:
    htag = h["tag"]
    htext = h["text"]
    indent = "  " * (int(htag[1]) - 1)
    print(f"{indent}{htag}: {htext}")
EOF

Step 6 -- Analyze interactive elements

openbrowser-ai -c - <<'EOF'
state = await browser.get_browser_state_summary()
elements_by_tag = {}
for idx, el in state.dom_state.selector_map.items():
    tag = el.tag_name
    elements_by_tag.setdefault(tag, []).append({
        "index": idx,
        "text": el.get_all_children_text(max_depth=1)[:50],
        "type": el.attributes.get("type", ""),
        "href": el.attributes.get("href", "")[:50] if el.attributes.get("href") else "",
    })

for tag, elems in sorted(elements_by_tag.items()):
    print(f"\n{tag} ({len(elems)} elements):")
    for e in elems[:5]:
        eidx = e["index"]
        etxt = e["text"]
        etype = e["type"]
        ehref = e["href"]
        print(f"  [{eidx}] text=\"{etxt}\" type={etype} href={ehref}")
    if len(elems) > 5:
        print(f"  ... and {len(elems) - 5} more")
EOF

Step 7 -- Page dimensions and scroll analysis

openbrowser-ai -c - <<'EOF'
dims = await evaluate("""
(function(){
  return {
    viewportWidth: window.innerWidth,
    viewportHeight: window.innerHeight,
    scrollHeight: document.body.scrollHeight,
    scrollWidth: document.body.scrollWidth,
    scrollable: document.body.scrollHeight > window.innerHeight
  };
})()
""")

import json
print(json.dumps(dims, indent=2))
if dims["scrollable"]:
    pages = dims["scrollHeight"] / dims["viewportHeight"]
    print(f"Page is approximately {pages:.1f} viewport heights long")
EOF

Step 8 -- Search for specific content patterns

openbrowser-ai -c - <<'EOF'
import re

# Get page text for Python-side analysis
text_content = await evaluate("document.body.innerText")

# Find emails
emails = re.findall(r"[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,}", text_content)
print(f"Emails found: {emails}")

# Find phone numbers
phones = re.findall(r"\+?\d[\d\s()-]{7,}", text_content)
print(f"Phone numbers found: {phones}")

# Find dates
dates = re.findall(r"\d{4}-\d{2}-\d{2}|\w+ \d{1,2},? \d{4}", text_content)
print(f"Dates found: {dates}")
EOF

Tips

  • Code is piped via stdin using heredoc (-c - <<'EOF'), so all Python syntax works without shell escaping issues.
  • Start with evaluate() for metadata and DOM statistics -- gives a fast structured overview.
  • Use browser.get_browser_state_summary() for interactive element analysis.
  • Use Python regex on extracted text for pattern matching (emails, phones, dates, prices).
  • For long pages, use await scroll(down=True) and re-extract to analyze below-fold content.
  • Variables persist between -c calls while the daemon is running, so you can build a comprehensive analysis incrementally.

Cleanup

This step is mandatory. Run it after the analysis finishes, whether extraction succeeded or the page failed to load. Without it, the daemon keeps Chrome running until its 10-minute idle timeout, leaving a stale browser process, a locked profile, and (on macOS/Linux desktop) a visible window.

Stop the daemon, then verify it is gone:

openbrowser-ai daemon stop
openbrowser-ai daemon status

daemon stop closes every tab, exits Chrome, flushes saved cookies/login state to the profile, and shuts down the daemon process. daemon status should report the daemon is not running. If it still reports running, run openbrowser-ai daemon doctor and then openbrowser-ai daemon reap. reap stops only verified OpenBrowser daemons and removes stale sockets and PID files.

If your invocation can fail mid-workflow (timeout, navigation error, malformed DOM), guarantee cleanup with a shell trap so the browser is never left orphaned:

trap 'openbrowser-ai daemon stop >/dev/null 2>&1 || true' EXIT
# ... openbrowser-ai -c calls here ...

Do not rely on the idle timeout. Do not call done() as a substitute, done() only marks the task complete inside the agent loop, it does not close the browser.

add-product-e2e-evalpaperclipai/paperclipAdd or extend a Paperclip full-stack runner E2E workflow, fixture, matcher, or report evidence path for local or Daytona execution.Stars: 114kUpdated Claude CodeCodexVerifiedInstall
add-runner-evalpaperclipai/paperclipAdd or extend a Paperclip Runner protocol evaluation definition, roster, assertion, or report fixture with provenance and narrow validation.Stars: 114kUpdated Claude CodeCodexVerifiedInstall
diagnose-why-work-stoppedpaperclipai/paperclipCompatibility entry point for failed, stalled, looping, or over-recovered Paperclip work. Use fix-failed-paperclip-run for the full diagnosis, regression, repair, real verification, and green-PR workflow; honor an explicitly requested diagnosis-only scope.Stars: 114kUpdated Claude CodeCodexVerifiedInstall
fix-failed-paperclip-runpaperclipai/paperclipDiagnose and fix failed, stuck, slow, or incorrectly blocked Paperclip tasks and runs. Use with a task/run URL or failure report to trace the root cause, prove a regression red then green, check second-order effects, verify the real workflow, and deliver a PR with passing checks.Stars: 114kUpdated Claude CodeCodexVerifiedInstall