Persist Browser tab ownership and URLs in the shared KVP store, migrate an existing Chromium session once, and restore tabs automatically for per-chat and shared scopes without showing Chromium's crash advisory.
Add composer-style microphone and arrow controls to the annotation dialog while preserving draft-versus-send voice behavior.
Keep dependency reconciliation, cache migration, and Chromium installation under the Browser hook's single setup lock. Remove the redundant nested installer lock from the Playwright helper.
Route every chat through one persistent Chromium, Xvfb, and Xpra runtime while keeping tab ownership and per-chat limits context-bound.
Share sign-ins across chats, adopt the first requesting legacy profile on upgrade, and close only the owning chat's tabs during reset or removal.
Keep separate and shared tab-strip modes as visibility choices and cover cross-context access, popups, screenshots, startup ownership, and cleanup.
Activate the selected Browser preset after the first Browser tool call so follow-up model turns use it.
Clear the temporary override at monologue end and preserve the main model as the safe fallback.
Render the existing Patchright page through an isolated, authenticated Xpra session with CDP screencast and snapshot fallbacks.
Keep Chromium out of fullscreen, synchronize native viewport resizing across canvas and modal handoffs, and remove Xpra decoration and shadow-cursor artifacts.
Move loading feedback into each Browser tab while reserving the footer for actual errors.
Run Chromium headful through Patchright on a private Xvfb display while keeping Agent Zero's page helpers in Patchright's isolated world.
Reconcile the pinned Patchright package and matching architecture-specific Chromium from Browser plugin hooks so fresh Docker images and self-updated installations converge on the same runtime.
Add project-scoped HTTP and SOCKS proxy settings for the internal Docker browser, including bypass and optional authentication. Pass them through Playwright's native proxy option, restart active Browser runtimes when they change, and update regression coverage and user documentation.
Return WebUI navigation after the initial document commit and align the socket timeout with browser command deadlines.
Remove forced viewport nudges and reject cropped or mismatched screencast frames so tab changes and pointer movement do not stretch the viewport.
Skip redundant same-version reinstalls and stage updates before replacing extension files.
Show clear install progress, offload blocking work, and restart active browser runtimes safely.
Make model presets a single global collection and treat presets as reusable setups (main, utility, embedding). Persisted scope configs now store only a preset name (Default is always present and immutable); project-scoped preset definitions were removed and attempts to edit/reset project presets are rejected. Add embedding-model support throughout: include embedding slot in presets, surface embedding in the switcher and override APIs, and notify extensions when the effective embedding model changes. Implement rename/retire propagation to update plugin configs, saved chats, and live AgentContext objects when preset names change or presets are removed. Update CLI/integration commands to use an "inherit" action for returning to the scoped preset and to report the effective scoped-or-chat preset. Add startup migration/bootstrap to initialize presets from remote or bundled fallback and numerous docs/UI text updates to reflect the new preset model. Extra: refactor config resolution helpers and add safety checks (validation, atomic writes) when saving presets.
Preserve explicit endpoint case, accept host:port and HTTP discovery addresses in Browser settings, and keep live-runtime guidance independent of any fixed localhost port.
Poll connected A0 CLI host-browser metadata while Browser Settings is open so newly authorized endpoints appear without saving or reopening.
Extend Browser Settings labels and remote-debugging guidance for Brave, Opera, Vivaldi, and Chromium-family host browsers with regression coverage.
Add host_browser_selection config normalization, Browser Settings UI choices from CLI-advertised inventory, and browser_selection forwarding through connector browser operations.
Expose host browser ids, labels, and available_browsers through A0 connector metadata and the browser_runtime API, with regression coverage for selection normalization.
Expose Browser settings for separate per-chat tabs versus a shared tab strip, and surface the existing maximum-tabs-per-chat cap. Make viewer WebSocket payloads carry the selected tab scope so the Browser panel can render the right tab list without guessing.
Keep the default scoped per chat, preserve the shared mode as an opt-in fallback, and cover both paths with Browser regression tests.
Notify screencast consumers when a stream is stopped so the WebSocket frame supervisor restarts after viewport-driven screencast resets.
Keep the active canvas as the frame/input surface and avoid replacing live screencasts with click snapshots, preserving hover, click, and wheel feedback.
Tested: conda run -n a0 pytest tests/test_browser_agent_regressions.py -q
Wake the WebSocket screencast supervisor when the runtime frame task stops before delivering a consumer frame, and keep fatal frame exits eligible for normal CDP teardown.
Avoid resizing the Browser canvas backing store on every frame when bitmap dimensions are unchanged, preventing unnecessary clears and reallocations.
Render live browser screencast frames through a canvas/ImageBitmap path while keeping the existing image/data-url path for snapshots and fallback rendering.
Attach WebSocket streams as runtime screencast consumers so delivered frames are pushed to the server loop and acknowledged after emit, removing the per-frame read_screencast_frame runtime round trip.
Negotiate binary and slim screencast frames so updated clients avoid base64-in-JSON and repeated viewer metadata on every frame.
Clamp screencast capture to the subscriber viewport and DPR, lower streaming JPEG quality while preserving snapshot quality, and teach the client to manage binary frame URLs safely.
Document the browser viewer performance plan and update regression coverage for the new transport contract.
Add oMLX as a first-class local chat and embedding provider using the Docker-friendly host.docker.internal:8000/v1 default and LiteLLM hosted_vllm wiring. Mark it as a no-key local provider, include it in onboarding with a bundled logo, document Apple Silicon setup including paged SSD cache, and cover the provider defaults, model search behavior, and onboarding metadata with focused tests.
Materialize browser, desktop, computer-use, and vision-load screenshots into chat-scoped artifacts so historical image refs survive temporary screenshot pruning.
Keep history serialization free of rescue assumptions, document durable screenshot behavior in tool prompts/skills, and size Xpra canvases from backend-normalized display dimensions to prevent stretched desktop views.
Verified with focused pytest coverage plus live Docker checks for browser screenshot persistence and Xpra canvas dimensions.
Add an Agent Zero owned browser DOM helper that captures shadow DOM and iframe content with frame-chain/node references.\n\nInstall the DOM helper before page-content capture for both local and host-browser runtimes, and send DOM helper payloads to A0 CLI host browser sessions when needed.\n\nCover iframe content refs and host-browser payload delivery in focused regression tests.
Preserve MCP image, audio, and resource tool results instead of collapsing non-text responses into an empty textual result. Images and image resources now flow into raw history as data URL attachments, while audio and non-image binary resources are saved as artifacts with normalized paths.
Extract shared media artifact helpers for base64 validation, image data URLs, decoded-size checks, artifact saving, MIME normalization, and safe filenames. Reuse the shared helpers from MCP, browser connector, and computer-use artifact paths, and add focused regression coverage.
Route no-path browser screenshots through an in-process ephemeral image registry that vision_load consumes into the existing data-url model boundary. Stop materializing host-browser artifacts into tmp/browser/host-screenshots, keep explicit path screenshots durable, and make browser log metadata point at the active chat/task context while preserving browser-context detail.
Deep-merge model preset slots with the active configuration so custom context windows, rate limits, and nested kwargs survive preset switches.
Treat legacy utility preset defaults as implicit values, allow omitted utility and embedding slots to inherit configured models, and document the partial-preset behavior.
Clarify Browser settings around internal Docker vs A0 CLI host-browser runtimes. Add recovery guidance to host-browser failures so users can switch back to the internal Docker browser from settings or /browser container. Cover the recovery messaging in host-browser connector tests.
Route host/local browser requests through the Browser tool instead of desktop or shell fallbacks. Add remote-debugging setup guidance to Browser runtime errors and document the exact Chrome inspect setting in prompts, skills, and Web UI copy.
Teach the Browser content helper to ignore global/delegated framework event bindings so snapshots surface the actual actionable controls instead of broad wrapper elements. Add an accessible name to the Browser address bar for more reliable capture output.
Allow agents to use selector-based reference actions, coordinate click fallbacks, focused-field typing, and string key chords such as CTRL+A across the browser tool, container runtime, and host connector runtime. Cover the behavior with browser regression and host connector tests.
Rename high-impact skills to task-oriented names and move plugin-owned skills into their owning plugin folders.\n\nAlign renamed skill frontmatter with the official SKILL.md standard by keeping trigger language in name/description metadata, replacing the old create-skill wizard with build-skill, and updating browser, A0 connector, computer-use, CLI setup, and scheduler skill references.\n\nTighten the recurring cross-provider guidance gaps surfaced by the evidence sweeps: memory requests now avoid promptinclude-file routing, scheduler prompts distinguish cron schedules from planned ISO dates, document questions prefer document_query, skills_tool search/read_file usage is clearer, normal notifications set info/priority 10, and local/host text editors preserve patch intent.\n\nUpdate regression tests for the renamed skills, plugin ownership, prompt budget reality, and standard frontmatter shape.
Default Bring Your Own Browser mode to the existing browser profile while exposing a clean Agent profile option in Browser settings with a clear warning for existing-profile access.
Forward the selected profile mode through the connector browser runtime, tolerate legacy config modules and old saved configs, and update regression coverage for the new payload shape.
Add a protected connector endpoint for reading and updating the Browser plugin runtime backend so the A0 CLI can switch between Docker browser and Bring Your Own Browser mode. Keep legacy host_when_available values normalized to host_required, move the host/container setting to the top of Browser settings, and cover the config normalization path.
Remove the ambiguous Use host when ready option from the Browser plugin settings and present the host-required path as Bring Your Own Browser. Add concise Chrome/Chromium remote-debugging guidance, normalize legacy host_when_available values to the BYOB setting, and make missing host-browser connector setup a repairable error with regression coverage.
Move URL normalization into Agent Zero-owned Browser helper code and expose the content helper's required API contract from the shared asset. Normalize host-browser open/navigate payloads before they cross into the connector, including nested multi actions, and add regression coverage for helper payload delivery and URL edge cases.
Make the Browser plugin the source of truth for browser-page-content.js by attaching its source and sha256 to host-browser operations when the CLI has no matching helper hash. Store the helper hash in connector metadata and cover the routing/ensure path in tests.
Store and surface host-browser preparation and CDP endpoint metadata from A0 CLI.
Let Browser runtime prepare candidate CLIs before the first action, and keep host-required errors more actionable.
Simplify Host Browser settings language and document the Chrome remote-debugging consent flow.
Integrate host-browser routing into the existing Browser tool. Store connector host-browser metadata, add pending browser op resolution, select connector runtimes from Browser settings, enforce host-content privacy policy, support automatic host preparation, and document the A0 CLI host-browser flow.
Use /a0/tmp/playwright as the Browser plugin Chromium cache and Docker install target while preserving full Chromium installs.
Add startup migration cleanup for retired usr Playwright caches, update Browser status/runtime references and docs, and cover migration behavior with focused regressions.
Introduce the shared surfaces frontend service and stylesheet so Browser and Desktop can register docked or floating live UI without special cases in modals.js. Update Browser and right-canvas integration to preserve active viewers across canvas/modal switches and avoid creating blank tabs unless explicitly requested.
Adds the explicit browser:screenshot action that writes JPEG/PNG files for vision_load, extends agent-callable Browser input actions, and documents the explicit vision workflow.
Adds the browser-forms on-demand skill and regression coverage for dispatch, runtime screenshot files, ref point resolution, upload path normalization, prompt discoverability, and label-wrapped form controls surfaced by the chat-driven E2E.
Detect cached Playwright contexts that have already closed before reusing the browser runtime.
Clear stale browser pages, popup waiters, screencasts, and interaction state; stop the old Playwright instance; and restart cleanly on next use. Add regression coverage for stale context recovery and unexpected context close events.
Detect openable Chrome extension UI pages from manifests and expose resolved chrome-extension URLs to the Browser UI.
Render an Open button in the compact Browser extension dropdown and cover manifest UI metadata with regression tests.
Bridge copy, cut, paste, and common edit shortcuts from the Browser modal and canvas screenshot surface into the Playwright runtime while preserving native clipboard behavior for Agent Zero UI fields.
Add websocket and runtime clipboard handling with regression coverage for frontend shortcut routing, paste fallback, and viewer input dispatch.
Expose extension deletion from the Browser internal settings page and keep the compact Browser dropdown focused on quick enable/install actions.\n\nAdd a guarded uninstall API that only deletes Browser-managed extension folders, updates enabled extension paths, refreshes the settings UI, and covers managed versus external paths with regression tests.
- Auto-register tabs opened by site (window.open, target=_blank,
ctrl-click) via context.on("page",...) with registry lock and
closing-state guard.
- Modifier-key click via Playwright trusted input: keyboard.down/up
around mouse.click for coord-based path; locator.click(modifiers=...)
selector fallback for off-screen / hidden elements. Chrome focus
rule: ctrl/meta-click keeps focus on origin tab; override via
focus_popup arg.
- key_chord action: presses keys in order, releases in reverse;
guarantees release on exception. Supports Ctrl+A/C/V style chords.
- mouse modifiers click-only (raises ValueError for non-click events).
- list(include_content=true) bulk read across all tabs in parallel
via asyncio.gather (was sequential).
- multi action: batched sub-calls. Different browser_id groups run
concurrently; same browser_id sequentially. Returns array of
{ok, result|error} matching input order. Lets the agent fan out
reads or coordinated mutations across tabs in one tool call.
- Cross-tab work no longer steals viewer focus.
last_interacted_browser_id promotes only on open / set_active /
same-tab work / Chrome popup rule. WebUI auto-open allowlist
tightened to open|navigate|set_active so background actions don't
drag the viewer.
- New set_active action for explicit focus switch.
- JS helper bumps VERSION to force re-injection on cached pages;
exports boundingBoxFor returning {x,y,w,h,selector} for the
trusted-input modifier-click paths.
Backwards-compatible: every new arg is optional with safe defaults.
No removed actions; existing call shapes preserved.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Keep browser sessions context-qualified so tabs from different chats can coexist without closing on context switches.
Create a real chat context when Browser launches from dashboard/no selected context, preserving agent handoff for that session.
Move chat context detail out of visible tab labels and into hover tooltips using only real chat names, with regression coverage for the updated lifecycle.
Restart the canvas screencast after page-changing commands and remount viewport metrics when starting or resizing streams so canvas scrolling stays smooth across first mount, new tabs, and navigation.
Move Browser JS off Alpine global store lookups and onto direct store imports, tighten modal/canvas handoff state, and keep annotations aligned with accepted viewport frames.
Improve Browser tab close ergonomics, allow Chromium native error pages to render without blocking the UI, include right-canvas tab polish, and expand regression coverage for these paths.
Decode browser frames before display and only render frames that match the active viewer viewport, avoiding stretched stale screencast images during startup and resize.
Keep rejecting mismatched CDP screencast frames on the backend, extend canvas viewport settling, and cover the behavior with browser regression tests.
Include small browser panel CSS polish.
Add Codex-inspired annotation UI to the built-in Browser surfaces, including the Annotate toggle, Cmd/Ctrl+. shortcut, selection overlay, inline comments, and batch Draft to chat / Send now actions.
Wire browser_viewer_annotation through the WebSocket and runtime layers, and expose safe DOM metadata extraction for clicked elements and selected areas without leaking password/value data.
Expand regression coverage for the Browser UI, annotation dispatch, runtime helper exposure, prompt formatting, and WebUI extension surface harness behavior.
