# Browser Research > Research web pages using agent-browser — a headless browser CLI that renders JavaScript and handles dynamic content. Use this skill as a fallback when WebSearch or WebFetch fails, returns insufficient results, or when the target page requires JS rendering (SPAs, dynamic docs). Also use when the user provides a specific URL to investigate, when you need to navigate multi-page documentation, or when summarizing web content that WebFetch cannot parse properly. Use when this capability is needed. - Skill: `tomevault-io/browser-research` (Agent Skill, multi-file: 2 files) - Install (CLI): `npx skillmds@latest add tomevault-io/browser-research` - Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/browser-research/raw - Safety review: pending (external: skill-scanner PASS, skillspector PASS) - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: AI & ML - Author: tomevault-io (https://skillmd.com/u/tomevault-io) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/tomevault-io/browser-research --- # Browser Research Skill Research web pages using agent-browser CLI and summarize content. ## Usage ```text /browser-research [research topic or question] ``` ## Workflow ### 0. Clean up existing session (if needed) A previous session may still be open. Close it before starting to avoid conflicts: ```bash agent-browser close 2>/dev/null || true ``` ### 1. Open the page ```bash agent-browser open "" && agent-browser wait --load networkidle --timeout 15000 ``` If `open` fails: verify the URL is well-formed, retry once. If it fails again, report the error to the user and stop. If `wait` times out: proceed anyway — the page may still be usable. **Next, choose the extraction method based on your purpose:** | Purpose | Command | When to use | |---------|---------|-------------| | Read article/docs text | `agent-browser eval "document.body.innerText"` | Blog posts, documentation, text-heavy pages. Most token-efficient | | Understand page structure | `agent-browser snapshot -c` | Need to see layout, navigation, or element refs for interaction | | Find interactive elements | `agent-browser snapshot -i -c` | Need to click links, buttons, or fill forms | For a typical single-article research, `eval "document.body.innerText"` is often sufficient. Use `snapshot` only when you need structure or element refs. **For large pages**, append `--max-output 10000` to prevent token explosion: ```bash agent-browser eval "document.body.innerText" --max-output 10000 ``` **If a cookie consent banner or overlay blocks content**, dismiss it first: ```bash agent-browser snapshot -i -c # find the accept/close button ref agent-browser click "@ref" # dismiss the banner ``` Then proceed with the chosen extraction method. **Stop here if you have enough information.** Steps 2–6 below are only needed for deeper investigation. ### 2. Get detailed content (if needed) Get text from specific element: ```bash agent-browser get text "@ref" ``` Get page metadata: ```bash agent-browser get title && agent-browser get url ``` Find elements by role, text, or label: ```bash agent-browser find role heading agent-browser find text "keyword" ``` ### 3. Handle long pages Scroll to load more content: ```bash agent-browser scroll down 500 && agent-browser snapshot -c ``` Scroll a specific element into view: ```bash agent-browser scrollintoview "@ref" && agent-browser snapshot -c ``` ### 4. Navigate to linked pages Click a link: ```bash agent-browser click "@ref" && agent-browser wait --load networkidle --timeout 15000 && agent-browser snapshot -c ``` Go back: ```bash agent-browser back && agent-browser snapshot -c ``` ### 5. Research additional URLs Use tabs to research multiple pages without losing previous context: ```bash agent-browser tab new && agent-browser open "" && agent-browser wait --load networkidle --timeout 15000 && agent-browser snapshot -c ``` Switch between tabs or close current tab: ```bash agent-browser tab list agent-browser tab agent-browser tab close ``` ### 6. Save page as PDF (optional) When the user requests a saved copy: ```bash agent-browser pdf "/path/to/output.pdf" ``` ### 7. Close when done ```bash agent-browser close ``` ## Critical Rules - **Always close the session** — every `open` must have a matching `close`. - **Read-only by default** — never submit forms or enter data. Clicking is allowed only for passive navigation: dismissing cookie/consent banners, following links, expanding collapsed sections, or switching tabs. Do not click buttons that trigger writes, purchases, or state changes. - **No guessing** — do not fabricate or assume page content; only report what `snapshot`/`get`/`eval` return. - **Authentication pages** — if a page requires login, report it immediately and stop. Do not attempt to authenticate. - **Prefer command chaining** — use `&&` to combine related commands in a single bash call for efficiency. - **Minimize tokens** — prefer `eval "document.body.innerText"` over `snapshot` when you only need text content. Use `--max-output` for large pages. ## Output Format Respond in the same language the user used. Summarize findings in this structure: 1. **Overview**: Main topic and purpose of the page 2. **Key Points**: Important information as bullet points 3. **Details**: Detailed explanations as needed 4. **Related Links**: Additional resources to reference When researching multiple URLs or when the user requests it, save results to a file using the Write tool. For a single-URL quick lookup, respond directly in chat. ## Snapshot Options | Flag | Description | |------|-------------| | `-i`, `--interactive` | Show only interactive elements | | `-c`, `--compact` | Remove empty structural elements | | `-d `, `--depth ` | Limit DOM tree depth | | `-s `, `--selector ` | Scope to CSS selector | ## Additional Useful Commands - `screenshot --full` — Capture full page screenshot - `screenshot --annotate` — Screenshot with numbered element labels - `diff snapshot` — Compare current page state against previous snapshot - `console` — View browser console logs (useful for debugging) - `errors` — View page errors - `get count ""` — Count matching elements - `--max-output ` — Truncate output for large pages --- > Converted and distributed by [TomeVault](https://tomevault.io/claim/kokatsu) — claim your Tome and manage your conversions.