Bohdan Triapitsyn
a5aa32446d
The preview panel worked by proxying a dev server through OpenChamber's own origin and rewriting the HTML that came back. Anything the rewriter did not anticipate broke, and pages that refuse to be embedded never loaded at all. This deletes the proxy (-1604 lines and its tests) and merges the preview and browser panels into one surface backed by a real Chromium view. What the panel is now - A `<webview>` in its own session partition: logins and cookies persist, hot reload works because nothing is rewritten, DevTools are one click away. - Annotation: pick one element, drag a region, or draw freehand, write a note, and it reaches chat with a screenshot of the visible page with the marks on it. - Toolbar: hard reload, page zoom, device sizes, a light/dark switch that applies to the page rather than the app, and cookie/cache clearing scoped to the panel alone. - Several pages at once, each tab showing the page's own favicon, and an address bar that suggests pages already visited in this project. - Dev servers are listed from what is actually listening on the machine, checked against what a project announced, so a server is offered no matter how it was started. One that is still starting is waited for instead of failing. Remote dev servers The desktop app binds a local port and pipes raw bytes to the OpenChamber host over the existing authenticated connection, so the page keeps its own origin at the root of its own host. The reachable set is exactly what discovery reports and is re-checked per connection, so an authenticated client cannot dial arbitrary local services on the host. Links and redirects to another loopback port stay on the machine that served the page. A tunnel that cannot be opened is reported; it is never replaced by the plain loopback URL, which would answer from the user's own machine under a remote address. Agent control Browser actions are a separate `openchamber_web` tool: open, snapshot, click, type, scroll, inspect computed styles, resize between mobile/tablet/desktop, and capture a screenshot into `.openchamber/screenshots/` in the project. The existing `openchamber` tool keeps sessions, worktrees and scheduled tasks. Each has its own setting in the new Settings -> General -> OpenChamber Tools section, and the plugin is not injected at all when both are off. Capability belongs to the connected client, not to configuration: a client declares on its event stream that it can drive a page, which only a Chromium host does. Exactly one client performs each request — it claims the request before acting, and the first claim wins — because deciding by whose result arrives first would be too late for a click that already happened. No client listening is answered immediately with an explanation rather than a timeout. Runtime boundaries Web tabs get a plain iframe that can display a page but not inspect one. The VS Code extension no longer offers the surface at all, since nothing that makes the panel worth having works there. Mobile is unaffected. Native boundary Camera, microphone, location and device-picker requests from panel pages are denied — Electron grants them by default when no handler is set, and the panel loads whatever address the user types. Page capture, appearance emulation and storage clearing verify that their target belongs to the panel's own session instead of trusting a web-contents id from the renderer. Persisted state Stored `preview` tabs migrate to `browser` (v13 -> v14). Context panel tab limits are now per surface, so filling one surface no longer evicts another's tabs. Address history is stored per project and per runtime. Documentation `preview.mdx` and `desktop-browser.mdx` rewritten across all locales, the agent tool settings path corrected, new `DOCUMENTATION.md` for the browser-control broker and the dev tunnel, and the `ui-api-decoupling` skill updated where it still described the deleted proxy.
2.8 KiB
2.8 KiB
Browser Control Broker
Purpose
This module carries agent browser actions from the server to the client that owns the in-app browser view, and the result back. The browser lives in a renderer, not in the server process, so the server can never act on a page itself; it can only ask and wait.
Boundaries
broker.jsowns request lifetime: it publishes one action through the injectedemitRequest, holds the pending request, and settles it on a client result, a timeout, or an abort signal. It knows nothing about transports.routes.jsis the result callback (POST /api/browser-control/result). It validates the envelope and hands the outcome to the broker.../../index.jssuppliesemitRequest, which writes the request to the OpenChamber SSE clients and returns how many were reached.../openchamber-control/service.jsis the only caller. It maps thebrowser.*actions of theopenchamber_webtool ontobroker.request()and owns their parameter validation.- The client half is
packages/ui/src/lib/browser/controlClient.ts, which registers the mounted browser pane as the one responder.
Invariants
- Capability belongs to the connection, not to configuration. A client declares
it can drive a page by opening its event stream with
browser=1, which only a Chromium host does; the flag lives and dies with that connection, so there is no setting to enable and no restart to remember. emitRequestcounts only clients that can serve the action.browser.openneeds any client, because opening a tab is what creates a view; every other action needs a declared-capable one.- Exactly one client performs a request. The broadcast reaches everyone who
could serve it, so a client claims the request over
POST /api/browser-control/claimand acts only if granted; the first claim wins and every other client does nothing. Deciding by whose result arrives first would be too late, because by then each of them has already clicked. A claim for a settled request is refused for the same reason. - Nobody listening is answered immediately with a 503 describing the environment, never by blocking for the full timeout. A blocked wait followed by a timeout cannot be told apart from a page that hung.
- A client that accepted a request and then disappeared still times out. Assuming success would report a page interaction that never happened.
- A result for an unknown request id is accepted with
matched: false, not an error: a client answering after the timeout has behaved correctly. - The result route parses its own body. This server has no global body parser, and a missing one silently turns every answer into an agent-visible timeout.
- Request payload limits are sized for a page snapshot (visible text plus every interactive element), not for a control message.