free-claude-code 6.3.3__tar.gz → 6.4.0__tar.gz

{free_claude_code-6.3.3 → free_claude_code-6.4.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: free-claude-code
-Version: 6.3.3
+Version: 6.4.0
 Summary: Local proxy connecting coding agents to OpenAI-compatible AI providers
 License-Expression: AGPL-3.0-only
 License-File: LICENSE
@@ -21,6 +21,7 @@ Requires-Dist: loguru>=0.7.0
 Requires-Dist: markdown-it-py>=4.2.0
 Requires-Dist: nvidia-riva-client>=2.26.0
 Requires-Dist: openai>=3.18.0
+Requires-Dist: packaging>=26.3
 Requires-Dist: pillow>=12.0.0; sys_platform == 'win32' or sys_platform == 'darwin'
 Requires-Dist: pydantic>=2.13.5
 Requires-Dist: pyperclip>=1.11.0
@@ -67,7 +68,7 @@ Description-Content-Type: text/markdown
 ## What You Get
 
 - **55 ToS-friendly providers. 1.3B+ free tokens every month.** Use free, paid, subscription, and local models from one searchable UI without putting your account at risk. FCC follows provider terms and removes integrations if they stop being allowed.
-- **10 coding agents. One model catalog.** Run [Claude Code](https://code.claude.com/docs/en/overview), [Codex](https://github.com/openai/codex), [Pi](https://github.com/earendil-works/pi), [OpenCode](https://github.com/anomalyco/opencode), [Cline](https://github.com/cline/cline), [Hermes](https://github.com/NousResearch/hermes-agent), [DeepSeek Harness](https://github.com/deepseek-ai/deepseek-harness), [Grok Build](https://github.com/xai-org/grok-build), [Muse Code](https://research.meta.ai/blog/introducing-muse-code-and-muse-spark-1-2/), or [Aider](https://aider.chat/) with your FCC models.
+- **11 coding agents. One model catalog.** Run [Claude Code](https://code.claude.com/docs/en/overview), [Codex](https://github.com/openai/codex), [Pi](https://github.com/earendil-works/pi), [OpenCode](https://github.com/anomalyco/opencode), [Cline](https://github.com/cline/cline), [Hermes](https://github.com/NousResearch/hermes-agent), [DeepSeek Harness](https://github.com/deepseek-ai/deepseek-harness), [Grok Build](https://github.com/xai-org/grok-build), [Muse Code](https://research.meta.ai/blog/introducing-muse-code-and-muse-spark-1-2/), [Aider](https://aider.chat/), or [VS Code Chat](https://code.visualstudio.com/) with your FCC models.
 - **Keep coding through provider outages.** After retries are exhausted, FCC automatically tries your next configured model without making you restart the turn. This works across every client.
 - **Up to 90% fewer terminal-output tokens.** Optional [RTK](https://github.com/rtk-ai/rtk) filters common command output, while five FCC optimizations handle quota probes, command-prefix detection, titles, suggestions, and filepaths without calling a provider.
 - **Native Code sessions in your browser.** Choose a folder and run Codex in the browser with real-time and background support. Freely switch providers/models in the same session. Support for switching harnesses in the same session coming soon!
@@ -400,6 +401,7 @@ For editor and app integrations, install the client, start FCC, then open
 **Admin UI → Integrations** and click **Connect** on its card.
 
 - **Claude Code in VS Code**: Install the [Claude Code extension](https://marketplace.visualstudio.com/items?itemName=anthropic.claude-code).
+- **VS Code Chat**: Install the latest [VS Code](https://code.visualstudio.com/), connect, and select an FCC model in its built-in Chat model picker.
 - **Claude Desktop**: Install [Claude Desktop](https://claude.ai/download). Fully quit it before connecting or disconnecting, then reopen it. Disconnect returns to normal Claude sign-in.
 - **Codex in VS Code and App**: Install the [Codex extension](https://marketplace.visualstudio.com/items?itemName=openai.chatgpt) or Codex App.
 - **Claude Code in JetBrains ACP**: Install Claude Agent in JetBrains AI Assistant and start it once, then click **Connect** in FCC. Reopen the IDE, select **Claude Code (FCC)**, and start a new chat. After JetBrains updates the agent, restart FCC before starting a new chat. Requires a local IDE in its standard installation locations.
@@ -503,6 +505,10 @@ Stop all running FCC commands, then run:
 fcc-update
 ```
 
+If FCC is already up to date, the command stops without running the installer. Otherwise, it shows the installed and available versions before updating. If the version check fails, retry the command.
+
+To reinstall FCC or change optional components while already up to date, run the [installer](#install) directly.
+
 For local voice support, include `--voice-local` (macOS/Linux) or `-VoiceLocal` (Windows), plus your `--torch-backend` or `-TorchBackend` option if used.
 
 If your installation does not have `fcc-update` yet, run the [installer](#install) once to add it.

{free_claude_code-6.3.3 → free_claude_code-6.4.0}/README.md

@@ -26,7 +26,7 @@
 ## What You Get
 
 - **55 ToS-friendly providers. 1.3B+ free tokens every month.** Use free, paid, subscription, and local models from one searchable UI without putting your account at risk. FCC follows provider terms and removes integrations if they stop being allowed.
-- **10 coding agents. One model catalog.** Run [Claude Code](https://code.claude.com/docs/en/overview), [Codex](https://github.com/openai/codex), [Pi](https://github.com/earendil-works/pi), [OpenCode](https://github.com/anomalyco/opencode), [Cline](https://github.com/cline/cline), [Hermes](https://github.com/NousResearch/hermes-agent), [DeepSeek Harness](https://github.com/deepseek-ai/deepseek-harness), [Grok Build](https://github.com/xai-org/grok-build), [Muse Code](https://research.meta.ai/blog/introducing-muse-code-and-muse-spark-1-2/), or [Aider](https://aider.chat/) with your FCC models.
+- **11 coding agents. One model catalog.** Run [Claude Code](https://code.claude.com/docs/en/overview), [Codex](https://github.com/openai/codex), [Pi](https://github.com/earendil-works/pi), [OpenCode](https://github.com/anomalyco/opencode), [Cline](https://github.com/cline/cline), [Hermes](https://github.com/NousResearch/hermes-agent), [DeepSeek Harness](https://github.com/deepseek-ai/deepseek-harness), [Grok Build](https://github.com/xai-org/grok-build), [Muse Code](https://research.meta.ai/blog/introducing-muse-code-and-muse-spark-1-2/), [Aider](https://aider.chat/), or [VS Code Chat](https://code.visualstudio.com/) with your FCC models.
 - **Keep coding through provider outages.** After retries are exhausted, FCC automatically tries your next configured model without making you restart the turn. This works across every client.
 - **Up to 90% fewer terminal-output tokens.** Optional [RTK](https://github.com/rtk-ai/rtk) filters common command output, while five FCC optimizations handle quota probes, command-prefix detection, titles, suggestions, and filepaths without calling a provider.
 - **Native Code sessions in your browser.** Choose a folder and run Codex in the browser with real-time and background support. Freely switch providers/models in the same session. Support for switching harnesses in the same session coming soon!
@@ -359,6 +359,7 @@ For editor and app integrations, install the client, start FCC, then open
 **Admin UI → Integrations** and click **Connect** on its card.
 
 - **Claude Code in VS Code**: Install the [Claude Code extension](https://marketplace.visualstudio.com/items?itemName=anthropic.claude-code).
+- **VS Code Chat**: Install the latest [VS Code](https://code.visualstudio.com/), connect, and select an FCC model in its built-in Chat model picker.
 - **Claude Desktop**: Install [Claude Desktop](https://claude.ai/download). Fully quit it before connecting or disconnecting, then reopen it. Disconnect returns to normal Claude sign-in.
 - **Codex in VS Code and App**: Install the [Codex extension](https://marketplace.visualstudio.com/items?itemName=openai.chatgpt) or Codex App.
 - **Claude Code in JetBrains ACP**: Install Claude Agent in JetBrains AI Assistant and start it once, then click **Connect** in FCC. Reopen the IDE, select **Claude Code (FCC)**, and start a new chat. After JetBrains updates the agent, restart FCC before starting a new chat. Requires a local IDE in its standard installation locations.
@@ -462,6 +463,10 @@ Stop all running FCC commands, then run:
 fcc-update
 ```
 
+If FCC is already up to date, the command stops without running the installer. Otherwise, it shows the installed and available versions before updating. If the version check fails, retry the command.
+
+To reinstall FCC or change optional components while already up to date, run the [installer](#install) directly.
+
 For local voice support, include `--voice-local` (macOS/Linux) or `-VoiceLocal` (Windows), plus your `--torch-backend` or `-TorchBackend` option if used.
 
 If your installation does not have `fcc-update` yet, run the [installer](#install) once to add it.

{free_claude_code-6.3.3 → free_claude_code-6.4.0}/e2e/conftest.py

@@ -1,7 +1,6 @@
 """Isolated browser-test composition for the local Admin UI."""
 
 import asyncio
-import faulthandler
 import socket
 import sys
 import threading
@@ -14,6 +13,7 @@ import uvicorn
 from playwright.sync_api import Page
 
 from e2e.code_support import CodeControl
+from e2e.server_shutdown import join_server
 from free_claude_code.api.app import create_app
 from free_claude_code.api.ports import ApiServices
 from free_claude_code.application.model_metadata import ProviderModelInfo
@@ -33,6 +33,7 @@ from free_claude_code.harnesses import (
     claude_integration,
     codex_integration,
     jetbrains_acp_integration,
+    vscode_chat_integration,
 )
 from free_claude_code.providers.base import BaseProvider, ProviderConfig
 from free_claude_code.providers.runtime import ProviderRuntime
@@ -185,6 +186,11 @@ def admin_base_url(
     monkeypatch.setattr(
         codex_integration, "config_path", lambda: tmp_path / ".codex" / "config.toml"
     )
+    monkeypatch.setattr(
+        vscode_chat_integration,
+        "config_path",
+        lambda: tmp_path / "vscode/chatLanguageModels.json",
+    )
     monkeypatch.setattr(
         claude_integration, "claude_state_path", lambda: tmp_path / ".claude.json"
     )
@@ -346,12 +352,9 @@ def admin_base_url(
         yield f"http://127.0.0.1:{port}"
     finally:
         server.should_exit = True
-        thread.join(timeout=5.0)
+        join_server(thread, code_control, request)
         listener.close()
         clear_settings_cache()
-        if thread.is_alive():
-            faulthandler.dump_traceback(file=sys.stderr)
-            pytest.fail("Admin browser-test server did not stop")
 
 
 @pytest.fixture(autouse=True)

free_claude_code-6.4.0/e2e/server_shutdown.py

@@ -0,0 +1,90 @@
+"""Bounded cleanup for the browser fixture's server thread."""
+
+import asyncio
+import faulthandler
+import inspect
+import sys
+import threading
+import time
+from concurrent.futures import Future
+
+import pytest
+
+from e2e.code_support import CodeControl
+
+
+def _task_snapshot(control: CodeControl) -> str:
+    jobs = control.service._jobs
+    lines = [f"Code jobs: {sum(not job.done() for job in jobs)}"]
+    for task in asyncio.all_tasks():
+        lines.append(f"Task {task.get_name()} (Code job: {task in jobs}):")
+        current = task.get_coro()
+        while inspect.iscoroutine(current) or inspect.isgenerator(current):
+            frame = (
+                current.cr_frame if inspect.iscoroutine(current) else current.gi_frame
+            )
+            if frame is not None:
+                lines.append(
+                    f"  {frame.f_code.co_filename}:{frame.f_lineno} in {frame.f_code.co_name}"
+                )
+            current = (
+                current.cr_await
+                if inspect.iscoroutine(current)
+                else current.gi_yieldfrom
+            )
+    return "\n".join(lines)
+
+
+def join_server(
+    thread: threading.Thread,
+    control: CodeControl,
+    request: pytest.FixtureRequest,
+    *,
+    diagnostic_after: float = 5,
+    timeout: float = 30,
+) -> None:
+    # Native interruption can itself take five seconds before durable cleanup.
+    # Diagnose at that point, but give the entire fixture shutdown its own budget.
+    started = time.monotonic()
+    deadline = started + timeout
+    thread.join(min(diagnostic_after, timeout))
+    if not thread.is_alive():
+        return
+
+    print(
+        f"Server shutdown still pending after {time.monotonic() - started:.2f}s: "
+        f"{request.node.nodeid}",
+        file=sys.stderr,
+    )
+    faulthandler.dump_traceback(file=sys.stderr)
+    snapshot: Future[str] = Future()
+
+    def capture() -> None:
+        try:
+            snapshot.set_result(_task_snapshot(control))
+        except Exception as error:
+            snapshot.set_result(f"Server task snapshot failed: {type(error).__name__}")
+
+    if control.loop is None:
+        snapshot.set_result("Server loop unavailable for task snapshot")
+    else:
+        try:
+            control.loop.call_soon_threadsafe(capture)
+        except RuntimeError:
+            snapshot.set_result("Server loop closed before task snapshot")
+
+    thread.join(max(0, deadline - time.monotonic()))
+    print(
+        snapshot.result()
+        if snapshot.done()
+        else "Server loop did not provide a task snapshot",
+        file=sys.stderr,
+    )
+    if thread.is_alive():
+        reason = (
+            f"Admin browser-test server did not stop within {timeout:g}s: "
+            f"{request.node.nodeid}"
+        )
+        request.session.shouldstop = reason
+        faulthandler.dump_traceback(file=sys.stderr)
+        pytest.fail(reason)

{free_claude_code-6.3.3 → free_claude_code-6.4.0}/e2e/test_admin_integrations.py

@@ -25,7 +25,7 @@ def test_connection_check_uses_disabled_loading_button(
 
     def hold_check(route):
         pending.append(route)
-        page.evaluate("window.integrationCheckIntercepted = true")
+        page.evaluate("window.integrationChecks = (window.integrationChecks || 0) + 1")
 
     page.route(f"**/admin/api/integrations/{integration}", hold_check)
     page.goto(f"{admin_base_url}/admin/integrations")
@@ -33,7 +33,6 @@ def test_connection_check_uses_disabled_loading_button(
     for visit in range(2):
         if visit:
             page.get_by_role("button", name="Providers", exact=True).click()
-            page.evaluate("window.integrationCheckIntercepted = false")
             page.get_by_role("button", name="Integrations", exact=True).click()
         expect(button).to_be_disabled()
         expect(button).to_have_text("Loading…")
@@ -46,9 +45,17 @@ def test_connection_check_uses_disabled_loading_button(
             == "integration-spinner"
         )
         expect(page.locator("#view-integrations .status-pill")).to_have_count(0)
-        page.wait_for_function("window.integrationCheckIntercepted === true")
+        page.wait_for_function(
+            "count => window.integrationChecks >= count", arg=visit + 1
+        )
         assert len(pending) == 1
-        pending.pop().fulfill(json={"connected": connected, "paths": None})
+        pending.pop().fulfill(
+            json={
+                "connected": connected,
+                "paths": None,
+                "update": {"state": "ready", "changed": False, "message": None},
+            }
+        )
         expect(button).to_be_enabled()
         expect(button).to_have_text("Disconnect" if connected else "Connect")
         expect(button).to_have_attribute("aria-busy", "false")
@@ -63,13 +70,10 @@ def test_codex_connect_disconnect_and_modal_paths(
     expect(page.locator("#openClaudeIntegration")).to_be_enabled()
     expect(page.locator("#messageArea")).to_have_text("")
     cards = page.locator("#view-integrations > article")
-    expect(cards).to_have_count(4)
+    expect(cards.first).to_be_visible()
     expect(page.locator("#claudeIntegrationStatus")).to_have_count(0)
     expect(page.locator("#openCodexIntegration")).to_be_enabled()
     expect(page.locator("#codexIntegrationStatus")).to_have_count(0)
-    expect(cards.nth(1)).to_contain_text(
-        "Use FCC's models in the Codex VS Code extension and desktop app."
-    )
     bounds = [card.bounding_box() for card in cards.all()]
     if width >= 1200:
         descriptions = [
@@ -124,8 +128,7 @@ def test_codex_connect_disconnect_and_modal_paths(
     expect(opener).to_have_css("color", "rgb(239, 68, 68)")
     if width >= 1200:
         buttons = [
-            button.bounding_box()
-            for button in page.locator(".integration-card > button").all()
+            card.get_by_role("button").first.bounding_box() for card in cards.all()
         ]
         for _, row in groupby(range(len(bounds)), key=lambda index: bounds[index]["y"]):
             assert len({buttons[index]["y"] for index in row}) == 1

{free_claude_code-6.3.3 → free_claude_code-6.4.0}/e2e/test_admin_providers.py

@@ -173,18 +173,19 @@ def test_delayed_startup_status_does_not_replace_a_manual_provider_check(
             route.continue_()
         else:
             startup.append(route)
+            page.evaluate("window.startupRequests = (window.startupRequests || 0) + 1")
+
+    def hold_manual(route: Route) -> None:
+        manual.append(route)
+        page.evaluate("window.manualRequests = (window.manualRequests || 0) + 1")
 
     page.route("**/admin/api/status", hold_first_status)
-    page.route(
-        "**/admin/api/providers/open_router/test", lambda route: manual.append(route)
-    )
-    page.expose_function("startupRequestIntercepted", lambda: bool(startup))
-    page.expose_function("manualRequestIntercepted", lambda: bool(manual))
+    page.route("**/admin/api/providers/open_router/test", hold_manual)
     page.evaluate("void refreshStartup()")
-    page.wait_for_function("window.startupRequestIntercepted()")
+    page.wait_for_function("window.startupRequests >= 1")
     dialog = open_provider(page, "open_router")
     dialog.get_by_role("button", name="Refresh models", exact=True).click()
-    page.wait_for_function("window.manualRequestIntercepted()")
+    page.wait_for_function("window.manualRequests >= 1")
     expected = "Checking..."
     if manual_result != "pending":
         manual.pop().fulfill(

free_claude_code-6.4.0/e2e/test_admin_vscode_chat.py

@@ -0,0 +1,92 @@
+import json
+
+import pytest
+from playwright.sync_api import expect
+
+from free_claude_code.harnesses import vscode_chat_integration as vscode
+
+
+def test_connect_and_disconnect_native_chat(page, admin_base_url):
+    path = vscode.config_path()
+    path.parent.mkdir(parents=True, exist_ok=True)
+    other = {"vendor": "customendpoint", "name": "Other", "models": []}
+    path.write_text(json.dumps([other]))
+    page.goto(f"{admin_base_url}/admin/integrations")
+    button = page.locator("#openVSCodeChatIntegration")
+    dialog = page.locator("#vscodeChatIntegrationDialog")
+    expect(button).to_have_text("Connect")
+    button.click()
+    expect(dialog).to_be_visible()
+    expect(dialog.locator("code")).to_have_text(str(path.resolve()))
+    page.keyboard.press("Escape")
+    expect(dialog).not_to_be_visible()
+    button.click()
+    dialog.get_by_role("button", name="Close", exact=True).click()
+    expect(dialog).not_to_be_visible()
+    button.click()
+    dialog.get_by_role("button", name="Connect", exact=True).click()
+    expect(button).to_have_text("Disconnect")
+    expect(button).to_have_class("danger-button")
+    group = json.loads(path.read_text())[1]
+    assert group["apiType"] == "messages"
+    assert group["models"]
+    assert all(m["url"].endswith("/v1/messages") for m in group["models"])
+    page.reload()
+    expect(button).to_have_text("Disconnect")
+    expect(page.locator("#vscodeChatIntegrationMessage")).not_to_be_visible()
+    button.click()
+    dialog.get_by_role("button", name="Disconnect", exact=True).click()
+    expect(button).to_have_text("Connect")
+    assert json.loads(path.read_text()) == [other]
+
+
+@pytest.mark.parametrize("status_read_failed", [False, True])
+def test_failed_update_has_actionable_retry(page, admin_base_url, status_read_failed):
+    progress: dict[str, str | bool | None] = {
+        "state": "failed",
+        "changed": False,
+        "message": "Could not update settings.",
+    }
+    retries = []
+
+    def startup(route):
+        response = route.fetch()
+        payload = response.json()
+        payload["startup"]["integrations"]["vscode-chat"] = dict(progress)
+        route.fulfill(response=response, json=payload)
+
+    def status(route):
+        if status_read_failed and progress["state"] == "failed":
+            route.fulfill(status=503, json={"detail": "Could not read settings."})
+        else:
+            route.fulfill(
+                json={"connected": True, "paths": None, "update": dict(progress)}
+            )
+
+    def retry(route):
+        retries.append(route.request.method)
+        progress.update(state="ready", message=None)
+        route.fulfill(json={"update": dict(progress)})
+
+    try:
+        page.route("**/admin/api/status", startup)
+        page.route("**/admin/api/integrations/vscode-chat", status)
+        page.route("**/admin/api/integrations/vscode-chat/refresh", retry)
+        page.goto(f"{admin_base_url}/admin/integrations")
+        main = page.locator("#openVSCodeChatIntegration")
+        secondary = page.locator("#retryVSCodeChatIntegration")
+        if status_read_failed:
+            expect(main).to_have_text("Retry")
+            action = main
+        else:
+            expect(main).to_have_text("Disconnect")
+            expect(secondary).to_be_visible()
+            action = secondary
+        with page.expect_response("**/admin/api/integrations/vscode-chat/refresh"):
+            action.click()
+        assert retries == ["POST"]
+        expect(main).to_have_text("Disconnect")
+        expect(secondary).to_be_hidden()
+        expect(page.locator("#vscodeChatIntegrationMessage")).to_be_hidden()
+    finally:
+        page.unroute_all(behavior="wait")

{free_claude_code-6.3.3 → free_claude_code-6.4.0}/e2e/test_integration_notices.py

@@ -4,6 +4,7 @@ from playwright.sync_api import expect
 
 @pytest.fixture(
     params=[
+        ("vscode-chat", "openVSCodeChatIntegration", "vscodeChatIntegrationMessage"),
         ("claude-vscode", "openClaudeIntegration", "claudeIntegrationMessage"),
         ("codex", "openCodexIntegration", "codexIntegrationMessage"),
         ("jetbrains-acp", "openJetBrainsIntegration", "jetBrainsIntegrationMessage"),
@@ -83,6 +84,68 @@ def test_notices_require_a_live_changed_completion(page, notice):
     expect(message).not_to_have_text("Update failed")
 
 
+def test_settled_poll_after_mutation_preserves_next_action(page, notice):
+    integration, button, message, observe = notice
+    observe("ready")
+    page.wait_for_function("!state.startupRequest && state.startupTimer === null")
+    unexpected_reads = []
+
+    def fail_status(route):
+        unexpected_reads.append(route.request)
+        route.fulfill(status=503, json={"detail": "Status unavailable"})
+
+    endpoint = f"**/admin/api/integrations/{integration}"
+    page.route(endpoint, fail_status)
+    dialog = page.get_by_role("dialog")
+    for action, connected in [("Disconnect", False), ("Connect", True)]:
+        page.route(
+            f"{endpoint}/{action.lower()}",
+            lambda route: route.fulfill(
+                json={
+                    "connected": route.request.url.endswith("/connect"),
+                    "disconnect_pending": False,
+                    "paths": None,
+                }
+            ),
+        )
+        button.click()
+        dialog.get_by_role("button", name=action, exact=True).click()
+        expect(dialog).not_to_be_visible()
+        next_action = "Disconnect" if connected else "Connect"
+        expect(button).to_have_text(next_action)
+        button.click()
+        observe("ready")
+        expect(
+            dialog.get_by_role("button", name=next_action, exact=True)
+        ).to_be_enabled()
+        assert not unexpected_reads
+        dialog.get_by_role("button", name="Close", exact=True).click()
+    page.unroute(endpoint, fail_status)
+    observe("failed")
+    expect(message).to_have_text("Update failed")
+    observe("ready")
+    expect(message).to_be_hidden()
+
+
+def test_missed_starting_state_still_reconciles_failure_and_recovery(page, notice):
+    integration, button, message, observe = notice
+    expect(message).to_be_hidden()
+    observe("failed")
+    expect(message).to_have_text("Update failed")
+    expect(button).to_be_enabled()
+    if integration == "vscode-chat":
+        expect(button).to_have_text("Disconnect")
+        expect(page.locator("#retryVSCodeChatIntegration")).to_be_visible()
+    observe("ready", changed=True)
+    expect(message).to_be_hidden()
+    expect(button).to_be_enabled()
+    if integration == "vscode-chat":
+        expect(page.locator("#retryVSCodeChatIntegration")).to_be_hidden()
+    # A repeated settled snapshot must not replay a reload notification.
+    observe("ready", changed=True)
+    expect(message).to_be_hidden()
+
+
 @pytest.mark.parametrize("revisit_while_busy", [False, True])
 def test_busy_status_recovery_preserves_notice_without_replaying_it(
     page, notice, revisit_while_busy
@@ -149,3 +212,31 @@ def test_new_server_and_off_page_completions_stay_quiet(page, notice):
     page.get_by_role("button", name="Integrations", exact=True).click()
     expect(button).to_be_enabled()
     expect(message).to_be_hidden()
+
+
+def test_missed_transition_during_status_read_is_reconciled(page, notice):
+    integration, button, message, observe = notice
+    pending = []
+
+    def hold_status(route):
+        pending.append(route)
+
+    pattern = f"**/admin/api/integrations/{integration}"
+    page.route(pattern, hold_status)
+    page.get_by_role("button", name="Providers", exact=True).click()
+    with page.expect_request(pattern):
+        page.get_by_role("button", name="Integrations", exact=True).click()
+    expect(button).to_be_disabled()
+    page.wait_for_function("!state.startupRequest && state.startupTimer === null")
+    observe("failed")
+    assert len(pending) == 1
+    page.unroute(pattern, hold_status)
+    pending[0].fulfill(
+        json={
+            "connected": True,
+            "paths": None,
+            "update": {"state": "ready", "changed": False, "message": None},
+        }
+    )
+    expect(message).to_have_text("Update failed")
+    expect(button).to_be_enabled()

{free_claude_code-6.3.3 → free_claude_code-6.4.0}/pyproject.toml

@@ -11,6 +11,7 @@ license = "AGPL-3.0-only"
 license-files = ["LICENSE"]
 requires-python = "==3.14.7"
 dependencies = [
+    "packaging>=26.3",
     "anyio>=4.15.1",
     "fastapi[standard]>=0.141.1",
     "uvicorn>=0.53.0",
@@ -41,6 +42,7 @@ dependencies = [
 ]
 
 [project.scripts]
+_fcc-update-check = "free_claude_code.updater.check:main"
 fcc-server = "free_claude_code.cli.entrypoints:serve"
 fcc-doctor = "free_claude_code.cli.entrypoints:doctor"
 fcc-claude = "free_claude_code.cli.launchers.claude:launch"
@@ -89,7 +91,6 @@ explicit = true
 
 [dependency-groups]
 dev = [
-    "packaging>=26.3",
     "pytest>=9.1.1",
     "pytest-asyncio>=1.4.0",
     "pytest-cov>=7.1.0",

{free_claude_code-6.3.3 → free_claude_code-6.4.0}/scripts/update/fcc-update

@@ -1,4 +1,11 @@
 #!/bin/sh
 set -eu
+status=0
+"$(dirname "$0")/_fcc-update-check" || status=$?
+case "$status" in
+    0) exit 0 ;;
+    10) ;;
+    *) exit "$status" ;;
+esac
 installer=$(curl -fsSL "https://raw.githubusercontent.com/Alishahryar1/free-claude-code/main/scripts/install.sh")
 exec sh -c "$installer" sh "$@"

{free_claude_code-6.3.3 → free_claude_code-6.4.0}/scripts/update/fcc-update.cmd

@@ -1,4 +1,7 @@
 @echo off
+"%~dp0_fcc-update-check.exe"
+if %errorlevel%==0 exit /b 0
+if not %errorlevel%==10 exit /b %errorlevel%
 (
   powershell.exe -NoProfile -ExecutionPolicy Bypass -Command "$ErrorActionPreference = 'Stop'; & ([scriptblock]::Create((Invoke-RestMethod 'https://raw.githubusercontent.com/Alishahryar1/free-claude-code/main/scripts/install.ps1')))" %*
   call exit /b %%errorlevel%%

{free_claude_code-6.3.3 → free_claude_code-6.4.0}/src/free_claude_code/api/admin_routes.py

@@ -273,6 +273,42 @@ async def claude_vscode_status(
     return await _integration_response(services.admin.claude_vscode_status)
 
 
+@router.get("/admin/api/integrations/vscode-chat")
+async def vscode_chat_status(
+    request: Request,
+    services: ApiServices = Depends(get_services),
+):
+    require_loopback_admin(request)
+    return await _integration_response(services.admin.vscode_chat_status)
+
+
+@router.post("/admin/api/integrations/vscode-chat/connect")
+async def connect_vscode_chat(
+    request: Request,
+    services: ApiServices = Depends(get_services),
+):
+    require_loopback_admin(request)
+    return await _integration_response(services.admin.connect_vscode_chat)
+
+
+@router.post("/admin/api/integrations/vscode-chat/disconnect")
+async def disconnect_vscode_chat(
+    request: Request,
+    services: ApiServices = Depends(get_services),
+):
+    require_loopback_admin(request)
+    return await _integration_response(services.admin.disconnect_vscode_chat)
+
+
+@router.post("/admin/api/integrations/vscode-chat/refresh")
+async def refresh_vscode_chat(
+    request: Request,
+    services: ApiServices = Depends(get_services),
+):
+    require_loopback_admin(request)
+    return await _integration_response(services.admin.refresh_vscode_chat)
+
+
 @router.post("/admin/api/integrations/claude-vscode/connect")
 async def connect_claude_vscode(
     request: Request,

{free_claude_code-6.3.3 → free_claude_code-6.4.0}/src/free_claude_code/api/admin_static/admin.css

@@ -266,19 +266,27 @@ textarea:focus-visible,
   align-items: flex-start;
   margin-bottom: 0;
 }
-.integration-card > button {
+.integration-actions {
+  display: flex;
+  gap: 8px;
+  margin-top: auto;
+}
+.integration-card > button,
+.integration-actions > button {
   display: inline-flex;
   align-items: center;
   gap: 8px;
   margin-top: auto;
 }
 .integration-card > button,
+.integration-actions > button,
 .provider-card > .provider-actions > button {
   min-height: 38px;
   padding: 6px 14px;
   font-size: 13px;
 }
 .integration-card > button[aria-busy="true"]::before,
+.integration-actions > button[aria-busy="true"]::before,
 button.startup-busy::before,
 .startup-spinner::before {
   content: "";
@@ -293,7 +301,7 @@ button.startup-busy::before,
   to { transform: rotate(360deg); }
 }
 @media (prefers-reduced-motion: reduce) {
-  .integration-card > button[aria-busy="true"]::before { animation: none; }
+  .integration-card button[aria-busy="true"]::before { animation: none; }
 }
 .integration-card h3 { min-height: 2lh; }
 .integration-dialog {