browser39

A headless web browser for AI agents. Converts pages to token-optimized Markdown locally. Single binary, no external browser.

browser39 fetches web pages and converts them to token-optimized Markdown that LLMs can actually consume. It runs JavaScript, manages cookies and sessions, queries the DOM, and fills forms. All processing happens locally, no data is sent to third-party services.

Works with: Claude Desktop & Claude Code | OpenClaw | Any agent via CLI

Comparison

Token savings in practice

Real test: extracting the "Optical communications" section from Artemis II on Wikipedia (full page: ~14,600 tokens).

browser39 returns the exact section in 196 tokens at zero cost. The raw approach misses it entirely, WebFetch burns 75x more tokens through an intermediate model, and cloud tools like Mistral's charge $0.03 per call.

Install

cargo install browser39

Pre-built binaries available on the releases page.

Auto-install prompts

Copy and paste into your agent to install browser39 automatically:

Claude Code

Install browser39 as an MCP server. Clone https://github.com/alejandroqh/browser39.git, build with cargo build --release, copy the binary to /usr/local/bin/browser39, then add it to MCP settings with command "browser39" and args ["mcp"].

OpenClaw

Install the browser39 plugin: openclaw plugins install https://github.com/alejandroqh/browser39.git && openclaw gateway restart

Quick Start

Claude Desktop / Claude Code (MCP)

Add to your MCP settings:

{
  "mcpServers": {
    "browser39": {
      "command": "browser39",
      "args": ["mcp"]
    }
  }
}

19 tools available instantly: browser39_fetch, browser39_click, browser39_links, browser39_dom_query, browser39_fill, browser39_submit, browser39_search, cookies, storage, history, and more.

See docs/install-claude.md for the full guide.

OpenClaw

openclaw plugins install https://github.com/alejandroqh/browser39.git
openclaw gateway restart

See docs/install-openclaw.md for bundle vs native plugin setup.

CLI: one-shot fetch

browser39 fetch https://example.com

# Example Domain
This domain is for use in documentation examples without needing permission.

[Learn more](https://iana.org/domains/example)

CLI: agent integration (watch mode)

Long-running subprocess that any language can talk to via JSONL files:

touch commands.jsonl
browser39 watch commands.jsonl --output results.jsonl

# From your agent (Python, Node, Rust, shell, anything):
echo '{"id":"1","action":"fetch","v":1,"seq":1,"url":"https://example.com"}' >> commands.jsonl

Drop-in web_search and visit_website tool examples: Python | TypeScript | Rust

See docs/install-cli.md for the full integration guide.

Features

Token optimization

browser39 minimizes token usage when feeding web content to LLMs:

• Content preselection: on first fetch, returns available content sections with token estimates instead of dumping the full page. The agent picks the relevant section and re-fetches with a targeted selector.
• Heading auto-expand: selector: "#Astronauts" returns the full section until the next same-level heading, not just the heading text.
• HTML to Markdown: strips scripts, styles, and non-content elements.
• Compact link references (JSON mode): [text][N] instead of inline URLs, with full URLs in the links array.
• Same-origin URL shortening: links on the same domain show path-only.
• Link deduplication: same-URL links (image + headline cards) emitted once.

JavaScript execution

boa_engine runs JS expressions against the parsed DOM. Supports document.querySelector, document.querySelectorAll, element properties, localStorage, document.cookie, form.submit(), and element.click().

{"action": "dom_query", "script": "document.querySelectorAll('a').length"}

Session persistence

Cookies, localStorage, and browsing history are persisted to disk by default (~/.local/share/browser39/session.enc, AES-256-GCM encrypted). An agent can log in once and stay authenticated across restarts.

Disable with --no-persist or config:

[session]
persistence = "memory"

Forms

Fill fields by CSS selector and submit. browser39 handles enctype, builds the HTTP request, and returns the response page:

{"action": "fill", "fields": [{"selector": "#user", "value": "agent"}, {"selector": "#pass", "value": "secret", "sensitive": true}]}
{"action": "submit", "selector": "form#login"}

Security

Auth profiles keep credentials out of the LLM conversation. The agent references a profile name and never sees the token:

[auth.github]
header = "Authorization"
value_env = "GITHUB_TOKEN"
value_prefix = "Bearer "
domains = ["api.github.com"]

{"action": "fetch", "url": "https://api.github.com/repos", "auth_profile": "github"}

All transports

Configuration

browser39 --config path/to/config.toml fetch https://example.com

Precedence: --config flag > BROWSER39_CONFIG env > ~/.config/browser39/config.toml

See docs/config.md for the full reference.

Documentation

Development

cargo build              # Build
cargo run                # Run
cargo test               # Run all tests
cargo clippy             # Lint
cargo fmt                # Format