Hypha Navigator

Let an AI agent drive your whole browser — over Hypha RPC. Tabs, navigation, DOM, screenshots, click/type by index, React, and arbitrary JavaScript on any page.

⬇ Download extension (.zip) v0.2.2 View on GitHub

MV3 · Chrome / Edge 116+ · load unpacked

What it does

🗂 Drive the browser

Open, close, switch, and navigate tabs across windows — list_tabs, open_tab, activate_tab, navigate, history.

🔎 Inspect & control pages

Smart indexed DOM, click/type by index, select, scroll, screenshots, query DOM, get HTML, and a React component tree.

⚡ Run JS anywhere

execute_script runs arbitrary JavaScript even on strict-CSP pages, via the Chrome debugger (CDP Page.setBypassCSP).

🧠 Per-site skills

The agent accumulates markdown Agent Skills bound to each site origin — it gets smarter and cheaper on every site over time.

📌 Stable target tab

Pin the tab the agent works on; keep browsing in other tabs. The pin survives the extension going to sleep.

📜 Self-documenting

Hand the agent one URL. GET /get_skill_md returns the full tool reference with curl examples.

Install

  1. Download the .zip above and unzip it somewhere permanent.
  2. Open chrome://extensions and enable Developer mode (top-right).
  3. Click Load unpacked and select the unzipped folder (the one with manifest.json).
  4. Pin the Hypha Navigator toolbar icon and click it to open the side panel.

Use it from an agent

Click Connect browser in the side panel, copy the Service URL, then:

curl "$SERVICE_URL/get_skill_md?_mode=last"          # full API + usage
curl "$SERVICE_URL/list_tabs?_mode=last"
curl -X POST "$SERVICE_URL/open_tab" -d '{"url":"https://example.com"}'
curl "$SERVICE_URL/get_browser_state?_mode=last"     # current target tab