diff --git a/PROGRESS.md b/PROGRESS.md
index 4d7c4b8..acd57b0 100644
--- a/PROGRESS.md
+++ b/PROGRESS.md
@@ -10,6 +10,44 @@ up instantly. Newest status at the top.
## ▶ Where we left off / next action
+- **Done:** **Screen-audio sharing wired into the Windows WinForms client** (2026-06-17).
+ The core already fully supported `SCREEN_AUDIO` capture on Windows (post-M3 WASAPI
+ loopback via `VOICECAT_HAS_LOOPBACK`, always on for the `windows-client` preset —
+ `core/CMakeLists.txt:85`; loopback start/stop at `client.cpp:1093`/`audio_engine.cpp:529`;
+ VAD/PTT/self-mute/server-mute correctly bypassed for non-MIC kinds at `client.cpp:862-879`)
+ and the C# Interop layer was already complete (`VcStreamKind.ScreenAudio`,
+ `StartStream`/`StopStream`/`SetRemoteStream`/`ListUserStreams` all generic). The gap was
+ purely UI wiring. **No core, proto, or C ABI changes were needed** — confirming the
+ "the core should support it already" assessment.
+ - `MainForm.Designer.cs` — new `btnScreenShareToggle` button in the voice panel top row
+ (`flpVoiceTop`), right after `btnMicToggle`, with full `AccessibleName`/
+ `AccessibleDescription` per the existing accessibility convention.
+ - `MainForm.cs` — new `_screenStreamId` field; `BtnScreenShareToggle_Click` handler
+ mirroring `BtnMicToggle_Click` but with no device picker / VAD / PTT / mode / mute
+ (screen audio bypasses all of those in the core). Independent of mic — can share
+ without joining voice and vice versa. `HandleDisconnected` now resets
+ `_screenStreamId` and disables the screen toggle. `OnFormClosed` now explicitly stops
+ both mic and screen streams before `Disconnect()` (clean `StreamStop` messages go out
+ before the control channel closes). `HandleStreamStarted` already labeled
+ `ScreenAudio => "screen audio"`; `PerUserTuningDialog.ApplySettings` already iterates
+ all of a peer's streams — both unchanged, peers can independently volume-tune a
+ screen-audio stream vs that user's mic.
+ - `VoiceCatClientSmokeTests.cs` — new `ScreenAudioStream_Starts_And_Stops` test:
+ connects + TOFU + guest auth, `StartStream(ScreenAudio)`, asserts
+ `VC_EVENT_STREAM_STARTED` arrives with matching `StreamId`, `StopStream`, asserts
+ `VC_EVENT_STREAM_STOPPED`. Passes headless (the `StreamAnnounce` succeeds regardless
+ of whether the loopback device initializes on a CI box).
+ - `dotnet build` — 0 warnings/errors. `dotnet test` — **4/4 tests green**
+ (3 existing + 1 new). Event trace confirms the full
+ `StreamStarted → UserUpdated → StreamStopped` path through P/Invoke against a live
+ `voicecat-server.exe`.
+ - **Not yet confirmed audible by ear** — pending manual two-instance live test (one
+ shares screen audio while something plays on the default render endpoint, the other
+ hears it). This is the observable-behavior exit criterion per `AGENTS.md`.
+ - **Documented caveat** (`docs/voice.md §9`, unchanged): whole-device WASAPI loopback
+ inherently re-captures this app's own incoming voice mix (self-echo loop) — accepted
+ characteristic, not a bug. Process-specific loopback (Windows 10 2004+
+ `AUDIOCLIENT_ACTIVATION_PARAMS`) is a future enhancement; miniaudio doesn't expose it.
- **In progress:** **M5 — moderation & admin** (2026-06-17). Server-side and C ABI are
implemented and tested: permissions, kick/ban/move/server-mute, channel CRUD, in-app account
management. Four new tests pass: `test_m5_permissions`, `test_m5_kick_ban_move_mute`,
diff --git a/clients/windows/VoiceCat.App/Forms/MainForm.Designer.cs b/clients/windows/VoiceCat.App/Forms/MainForm.Designer.cs
index 4583297..e1a96c9 100644
--- a/clients/windows/VoiceCat.App/Forms/MainForm.Designer.cs
+++ b/clients/windows/VoiceCat.App/Forms/MainForm.Designer.cs
@@ -36,6 +36,7 @@ partial class MainForm
private FlowLayoutPanel flpVoiceTop = null!;
private FlowLayoutPanel flpVoiceBottom = null!;
private Button btnMicToggle = null!;
+ private Button btnScreenShareToggle = null!;
private CheckBox chkMute = null!;
private CheckBox chkDeafen = null!;
private RadioButton radioVad = null!;
@@ -81,6 +82,7 @@ partial class MainForm
flpVoiceTop = new FlowLayoutPanel();
flpVoiceBottom = new FlowLayoutPanel();
btnMicToggle = new Button();
+ btnScreenShareToggle = new Button();
chkMute = new CheckBox();
chkDeafen = new CheckBox();
radioVad = new RadioButton();
@@ -227,6 +229,20 @@ partial class MainForm
btnMicToggle.Margin = new Padding(0, 2, 6, 0);
btnMicToggle.TabIndex = 0;
+ // Screen-audio share — independent of mic voice (can share without joining voice and
+ // vice versa). The core's WASAPI loopback path (VOICECAT_HAS_LOOPBACK, always on for
+ // the windows-client preset) captures the default render endpoint; see docs/voice.md
+ // §9. Whole-device loopback inherently re-captures this app's own incoming voice mix —
+ // an accepted self-echo characteristic, not a bug.
+ btnScreenShareToggle.Text = "Share Screen &Audio";
+ btnScreenShareToggle.AccessibleName = "Share screen audio";
+ btnScreenShareToggle.AccessibleDescription =
+ "Start or stop sharing your computer's audio (desktop/system audio) with the channel. " +
+ "Independent of the microphone. Captures everything playing through your default speakers.";
+ btnScreenShareToggle.AutoSize = true;
+ btnScreenShareToggle.Margin = new Padding(0, 2, 12, 0);
+ btnScreenShareToggle.TabIndex = 10;
+
chkMute.Text = "&Mute mic";
chkMute.AutoSize = true;
chkMute.Enabled = false;
@@ -276,6 +292,7 @@ partial class MainForm
flpVoiceTop.AutoSize = false;
flpVoiceTop.Padding = new Padding(4, 2, 4, 0);
flpVoiceTop.Controls.Add(btnMicToggle);
+ flpVoiceTop.Controls.Add(btnScreenShareToggle);
flpVoiceTop.Controls.Add(chkMute);
flpVoiceTop.Controls.Add(chkDeafen);
flpVoiceTop.Controls.Add(lblMode);
diff --git a/clients/windows/VoiceCat.App/Forms/MainForm.cs b/clients/windows/VoiceCat.App/Forms/MainForm.cs
index 0e6272c..24e69d8 100644
--- a/clients/windows/VoiceCat.App/Forms/MainForm.cs
+++ b/clients/windows/VoiceCat.App/Forms/MainForm.cs
@@ -23,6 +23,7 @@ public partial class MainForm : Form
// Voice state
private uint _micStreamId; // 0 = not started
+ private uint _screenStreamId; // 0 = not sharing screen audio
private Keys _pttKey = Keys.F8;
private bool _serverMuted;
private bool _serverDeafened;
@@ -60,6 +61,7 @@ public partial class MainForm : Form
// Voice controls
btnMicToggle.Click += BtnMicToggle_Click;
+ btnScreenShareToggle.Click += BtnScreenShareToggle_Click;
chkMute.CheckedChanged += (_, _) => ApplySelfMute();
chkDeafen.CheckedChanged += (_, _) => ApplySelfMute();
radioVad.CheckedChanged += RadioVad_CheckedChanged;
@@ -404,9 +406,11 @@ public partial class MainForm : Form
_talkingUsers.Clear();
_currentChannelId = 0;
_micStreamId = 0;
+ _screenStreamId = 0;
txtCompose.Enabled = false;
btnSend.Enabled = false;
btnMicToggle.Enabled = false;
+ btnScreenShareToggle.Enabled = false;
}
// ── Level meter ───────────────────────────────────────────────────────────
@@ -604,6 +608,35 @@ public partial class MainForm : Form
private void ApplySelfMute() =>
_client.SetSelfMute(chkMute.Checked, chkDeafen.Checked);
+ // Screen-audio share — independent of mic voice. The core bypasses VAD/PTT/self-mute/
+ // server-mute for non-MIC kinds (client.cpp on_capture_frame), so no input-mode or mute
+ // state applies here. WASAPI loopback captures the default render endpoint (whole-device,
+ // not process-specific — docs/voice.md §9).
+ private void BtnScreenShareToggle_Click(object? sender, EventArgs e)
+ {
+ if (_screenStreamId == 0)
+ {
+ var (result, streamId) = _client.StartStream(VcStreamKind.ScreenAudio, "Desktop audio");
+ if (result == VcResult.Ok)
+ {
+ _screenStreamId = streamId;
+ btnScreenShareToggle.Text = "Stop Screen &Audio";
+ AddActivity("Started sharing screen audio");
+ }
+ else
+ {
+ AddActivity($"Failed to start screen audio: {result}");
+ }
+ }
+ else
+ {
+ _client.StopStream(_screenStreamId);
+ _screenStreamId = 0;
+ btnScreenShareToggle.Text = "Share Screen &Audio";
+ AddActivity("Stopped sharing screen audio");
+ }
+ }
+
private void RadioVad_CheckedChanged(object? sender, EventArgs e)
{
if (!radioVad.Checked) return;
@@ -916,6 +949,11 @@ public partial class MainForm : Form
_pumpTimer.Stop();
_client.LevelChanged -= OnLevelChanged;
_client.EventReceived -= OnEvent;
+ // Stop any active local streams before tearing down — the core stops loopback in
+ // stream_stop/destroy, but explicit stops ensure clean StreamStop protocol messages
+ // go out before Disconnect closes the control channel.
+ if (_screenStreamId != 0) _client.StopStream(_screenStreamId);
+ if (_micStreamId != 0) _client.StopStream(_micStreamId);
_client.Disconnect();
_client.Dispose();
base.OnFormClosed(e);
diff --git a/clients/windows/VoiceCat.Interop.Tests/VoiceCatClientSmokeTests.cs b/clients/windows/VoiceCat.Interop.Tests/VoiceCatClientSmokeTests.cs
index 0cc95a1..4d4c0f4 100644
--- a/clients/windows/VoiceCat.Interop.Tests/VoiceCatClientSmokeTests.cs
+++ b/clients/windows/VoiceCat.Interop.Tests/VoiceCatClientSmokeTests.cs
@@ -246,4 +246,51 @@ public sealed class VoiceCatClientSmokeTests : IDisposable
client.Disconnect();
}
+
+ ///
+ /// Screen-audio (SCREEN_AUDIO) stream start/stop through the P/Invoke layer. The core's
+ /// WASAPI loopback path (VOICECAT_HAS_LOOPBACK) captures the default render endpoint; the
+ /// StreamAnnounce succeeds regardless of whether the loopback device actually initializes
+ /// on a headless box, so this test passes in CI while still exercising the full
+ /// StartStream -> StreamStarted -> StopStream -> StreamStopped path through P/Invoke.
+ /// See docs/voice.md §9 and MainForm's BtnScreenShareToggle_Click.
+ ///
+ [Fact]
+ public void ScreenAudioStream_Starts_And_Stops()
+ {
+ var events = new List();
+ using var client = new VoiceCatClient("vc-csharp-screen", "0.1", VcLogLevel.Off,
+ tofuStorePath: Path.Combine(_tempDir, "tofu_pins_screen.txt"));
+ client.EventReceived += events.Add;
+
+ Assert.Equal(VcResult.Ok, client.Connect("127.0.0.1", _port));
+ Assert.Equal(VcResult.Ok, client.AuthenticateGuest("CSharpScreen"));
+
+ Assert.True(PumpUntil(client, () => events.Any(e => e.Type == VcEventType.ServerIdentity), 5000));
+ Assert.Equal(VcResult.Ok, client.ConfirmServerIdentity(accept: true));
+ Assert.True(PumpUntil(client, () => events.Any(e => e.Type == VcEventType.AuthResult), 5000));
+ Assert.Equal(VcResult.Ok, events.First(e => e.Type == VcEventType.AuthResult).Result);
+ Assert.True(PumpUntil(client, () => events.Any(e => e.Type == VcEventType.ChannelList), 3000));
+
+ // Give the async UDP binding handshake a moment to land before announcing a stream
+ // (mirrors vccli's 500ms sleep after auth).
+ Thread.Sleep(500);
+
+ var (startResult, streamId) = client.StartStream(VcStreamKind.ScreenAudio, "Desktop audio");
+ Assert.Equal(VcResult.Ok, startResult);
+ Assert.True(streamId != 0, "StreamId should be non-zero on success");
+
+ // The core emits VC_EVENT_STREAM_STARTED for the local client too (client.cpp
+ // handle_stream_announce_result), so we see our own screen-audio stream start.
+ Assert.True(PumpUntil(client,
+ () => events.Any(e => e.Type == VcEventType.StreamStarted && e.StreamId == streamId), 5000),
+ "did not receive VC_EVENT_STREAM_STARTED for screen-audio stream");
+
+ Assert.Equal(VcResult.Ok, client.StopStream(streamId));
+ Assert.True(PumpUntil(client,
+ () => events.Any(e => e.Type == VcEventType.StreamStopped && e.StreamId == streamId), 5000),
+ "did not receive VC_EVENT_STREAM_STOPPED for screen-audio stream");
+
+ client.Disconnect();
+ }
}