agentvibes 5.13.1 → 5.14.0

package/.claude/hooks-windows/play-tts.ps1

@@ -509,6 +509,23 @@ if ($OverrideEffects -ne "" -and $OverrideEffects -in @("off", "light", "medium"
     }
 }
 
+# Fail loudly when mixing was requested but ffmpeg is unavailable. Without this,
+# background music and reverb are dropped silently and the voice plays "fine" —
+# indistinguishable from a config problem. A long-running tts-watcher.ps1 that
+# started before ffmpeg was installed keeps a stale PATH snapshot and hits this,
+# so the message names the restart explicitly. Mirrors the bash-side warning in
+# audio-processor.sh ("ffmpeg not installed, skipping background mix").
+if (-not $HasFfmpeg -and ($BgEnabled -or $HasReverb)) {
+    $_dropped = @()
+    if ($BgEnabled) { $_dropped += "background music" }
+    if ($HasReverb) { $_dropped += "reverb ($ReverbLevel)" }
+    Write-Host ("[WARNING] play-tts.ps1: ffmpeg not found on PATH — dropping " +
+        ($_dropped -join " and ") + "; voice will play unmixed.") -ForegroundColor Yellow
+    Write-Host ("[WARNING] Install ffmpeg (e.g. 'scoop install ffmpeg'). If it IS " +
+        "installed, a background tts-watcher started before it will have a stale " +
+        "PATH — restart the watcher.") -ForegroundColor Yellow
+}
+
 # --- Apply LLM-specific engine override --------------------------------------
 # Allowed local Windows engines: windows-sapi, windows-piper, soprano.
 # Transport providers (ssh-remote etc.) are not listed because they forward

package/.claude/hooks-windows/soprano-gradio-synth.py

@@ -1,153 +1,153 @@
-#!/usr/bin/env python3
-#
-# File: .claude/hooks-windows/soprano-gradio-synth.py
-#
-# AgentVibes - Finally, your AI Agents can Talk Back!
-# Website: https://agentvibes.org
-# Repository: https://github.com/paulpreibisch/AgentVibes
-#
-# Co-created by Paul Preibisch with Claude AI
-# Copyright (c) 2025 Paul Preibisch
-#
-# Licensed under the Apache License, Version 2.0 (the "License");
-# you may not use this file except in compliance with the License.
-# You may obtain a copy of the License at
-#
-#     http://www.apache.org/licenses/LICENSE-2.0
-#
-# Unless required by applicable law or agreed to in writing, software
-# distributed under the License is distributed on an "AS IS" BASIS,
-# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-# See the License for the specific language governing permissions and
-# limitations under the License.
-#
-# ---
-#
-# Soprano Gradio WebUI synthesizer helper.
-# Calls the Soprano WebUI's Gradio API and saves the result as a WAV file.
-# Uses only Python stdlib (json, sys, urllib) — no extra dependencies.
-#
-# Usage: python3 soprano-gradio-synth.py "text to speak" output.wav [port]
-#
-"""
-Soprano Gradio WebUI synthesizer helper for AgentVibes.
-
-Calls the Soprano WebUI's Gradio API via the Server-Sent Events (SSE) protocol
-and downloads the generated audio as a WAV file.
-
-Flow:
-  1. Submit generation request -> get event_id
-  2. Poll SSE stream for audio file URL
-  3. Download WAV file to output path
-
-See: https://github.com/ekwek1/soprano
-"""
-import json
-import sys
-import urllib.request
-import urllib.error
-
-
-def synth(text: str, output_path: str, port: int = 7860) -> None:
-    base = f"http://127.0.0.1:{port}"
-
-    # Step 1: Submit generation request
-    # Args: text, temperature, top_p, repetition_penalty, chunk_size, streaming
-    payload = json.dumps({
-        "data": [text, 0.0, 0.95, 1.2, 1, False]
-    }).encode()
-
-    event_id = submit_request(base, payload)
-
-    # Step 2: Poll SSE stream for audio file URL
-    audio_url = poll_for_result(base, event_id)
-
-    # Step 3: Download the audio file (validate localhost to prevent SSRF)
-    download_file(audio_url, output_path, port)
-
-
-def submit_request(base: str, payload: bytes) -> str:
-    """Submit generation request to Gradio API, return event_id."""
-    for api_base in ["/gradio_api/call", "/call"]:
-        url = f"{base}{api_base}/generate_speech"
-        req = urllib.request.Request(
-            url,
-            data=payload,
-            headers={"Content-Type": "application/json"},
-        )
-        try:
-            with urllib.request.urlopen(req, timeout=30) as resp:
-                return json.loads(resp.read())["event_id"]
-        except urllib.error.URLError:
-            continue
-
-    raise RuntimeError("Could not reach Soprano WebUI API")
-
-
-def poll_for_result(base: str, event_id: str) -> str:
-    """Poll SSE endpoint until audio file URL is returned."""
-    for api_base in ["/gradio_api/call", "/call"]:
-        url = f"{base}{api_base}/generate_speech/{event_id}"
-        req = urllib.request.Request(url)
-        try:
-            with urllib.request.urlopen(req, timeout=120) as resp:
-                for raw_line in resp:
-                    line = raw_line.decode("utf-8", errors="replace").strip()
-                    if not line.startswith("data: "):
-                        continue
-                    try:
-                        parsed = json.loads(line[6:])
-                    except json.JSONDecodeError:
-                        continue
-                    # Response format: [{"path": "...", "url": "...", ...}, "status string"]
-                    if isinstance(parsed, list) and len(parsed) >= 1:
-                        audio = parsed[0]
-                        if isinstance(audio, dict) and "url" in audio:
-                            return audio["url"]
-            break
-        except urllib.error.URLError:
-            continue
-
-    raise RuntimeError("No audio URL in Soprano response")
-
-
-def download_file(url: str, output_path: str, port: int = 7860) -> None:
-    """Download audio file from Gradio file server.
-
-    Security: Validates URL points to localhost to prevent SSRF.
-    """
-    # Validate URL is localhost to prevent SSRF via crafted Gradio responses
-    allowed_prefixes = [
-        f"http://127.0.0.1:{port}/",
-        f"http://localhost:{port}/",
-    ]
-    if not any(url.startswith(prefix) for prefix in allowed_prefixes):
-        raise RuntimeError(f"Refused download from non-localhost URL: {url[:80]}")
-
-    req = urllib.request.Request(url)
-    with urllib.request.urlopen(req, timeout=30) as resp:
-        with open(output_path, "wb") as f:
-            while True:
-                chunk = resp.read(8192)
-                if not chunk:
-                    break
-                f.write(chunk)
-
-
-if __name__ == "__main__":
-    if len(sys.argv) < 3:
-        print(f"Usage: {sys.argv[0]} \"text\" output.wav [port]", file=sys.stderr)
-        sys.exit(1)
-
-    text = sys.argv[1]
-    output = sys.argv[2]
-    port = int(sys.argv[3]) if len(sys.argv) > 3 else 7860
-    if not (1 <= port <= 65535):
-        print(f"Error: Invalid port number: {port}", file=sys.stderr)
-        sys.exit(1)
-
-    try:
-        synth(text, output, port)
-    except Exception as e:
-        print(f"Error: {e}", file=sys.stderr)
-        sys.exit(1)
+#!/usr/bin/env python3
+#
+# File: .claude/hooks-windows/soprano-gradio-synth.py
+#
+# AgentVibes - Finally, your AI Agents can Talk Back!
+# Website: https://agentvibes.org
+# Repository: https://github.com/paulpreibisch/AgentVibes
+#
+# Co-created by Paul Preibisch with Claude AI
+# Copyright (c) 2025 Paul Preibisch
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+#
+# ---
+#
+# Soprano Gradio WebUI synthesizer helper.
+# Calls the Soprano WebUI's Gradio API and saves the result as a WAV file.
+# Uses only Python stdlib (json, sys, urllib) — no extra dependencies.
+#
+# Usage: python3 soprano-gradio-synth.py "text to speak" output.wav [port]
+#
+"""
+Soprano Gradio WebUI synthesizer helper for AgentVibes.
+
+Calls the Soprano WebUI's Gradio API via the Server-Sent Events (SSE) protocol
+and downloads the generated audio as a WAV file.
+
+Flow:
+  1. Submit generation request -> get event_id
+  2. Poll SSE stream for audio file URL
+  3. Download WAV file to output path
+
+See: https://github.com/ekwek1/soprano
+"""
+import json
+import sys
+import urllib.request
+import urllib.error
+
+
+def synth(text: str, output_path: str, port: int = 7860) -> None:
+    base = f"http://127.0.0.1:{port}"
+
+    # Step 1: Submit generation request
+    # Args: text, temperature, top_p, repetition_penalty, chunk_size, streaming
+    payload = json.dumps({
+        "data": [text, 0.0, 0.95, 1.2, 1, False]
+    }).encode()
+
+    event_id = submit_request(base, payload)
+
+    # Step 2: Poll SSE stream for audio file URL
+    audio_url = poll_for_result(base, event_id)
+
+    # Step 3: Download the audio file (validate localhost to prevent SSRF)
+    download_file(audio_url, output_path, port)
+
+
+def submit_request(base: str, payload: bytes) -> str:
+    """Submit generation request to Gradio API, return event_id."""
+    for api_base in ["/gradio_api/call", "/call"]:
+        url = f"{base}{api_base}/generate_speech"
+        req = urllib.request.Request(
+            url,
+            data=payload,
+            headers={"Content-Type": "application/json"},
+        )
+        try:
+            with urllib.request.urlopen(req, timeout=30) as resp:
+                return json.loads(resp.read())["event_id"]
+        except urllib.error.URLError:
+            continue
+
+    raise RuntimeError("Could not reach Soprano WebUI API")
+
+
+def poll_for_result(base: str, event_id: str) -> str:
+    """Poll SSE endpoint until audio file URL is returned."""
+    for api_base in ["/gradio_api/call", "/call"]:
+        url = f"{base}{api_base}/generate_speech/{event_id}"
+        req = urllib.request.Request(url)
+        try:
+            with urllib.request.urlopen(req, timeout=120) as resp:
+                for raw_line in resp:
+                    line = raw_line.decode("utf-8", errors="replace").strip()
+                    if not line.startswith("data: "):
+                        continue
+                    try:
+                        parsed = json.loads(line[6:])
+                    except json.JSONDecodeError:
+                        continue
+                    # Response format: [{"path": "...", "url": "...", ...}, "status string"]
+                    if isinstance(parsed, list) and len(parsed) >= 1:
+                        audio = parsed[0]
+                        if isinstance(audio, dict) and "url" in audio:
+                            return audio["url"]
+            break
+        except urllib.error.URLError:
+            continue
+
+    raise RuntimeError("No audio URL in Soprano response")
+
+
+def download_file(url: str, output_path: str, port: int = 7860) -> None:
+    """Download audio file from Gradio file server.
+
+    Security: Validates URL points to localhost to prevent SSRF.
+    """
+    # Validate URL is localhost to prevent SSRF via crafted Gradio responses
+    allowed_prefixes = [
+        f"http://127.0.0.1:{port}/",
+        f"http://localhost:{port}/",
+    ]
+    if not any(url.startswith(prefix) for prefix in allowed_prefixes):
+        raise RuntimeError(f"Refused download from non-localhost URL: {url[:80]}")
+
+    req = urllib.request.Request(url)
+    with urllib.request.urlopen(req, timeout=30) as resp:
+        with open(output_path, "wb") as f:
+            while True:
+                chunk = resp.read(8192)
+                if not chunk:
+                    break
+                f.write(chunk)
+
+
+if __name__ == "__main__":
+    if len(sys.argv) < 3:
+        print(f"Usage: {sys.argv[0]} \"text\" output.wav [port]", file=sys.stderr)
+        sys.exit(1)
+
+    text = sys.argv[1]
+    output = sys.argv[2]
+    port = int(sys.argv[3]) if len(sys.argv) > 3 else 7860
+    if not (1 <= port <= 65535):
+        print(f"Error: Invalid port number: {port}", file=sys.stderr)
+        sys.exit(1)
+
+    try:
+        synth(text, output, port)
+    except Exception as e:
+        print(f"Error: {e}", file=sys.stderr)
+        sys.exit(1)

package/.claude/verbosity.txt

@@ -1 +1 @@
-high
+high

package/.clawdbot/README.md

@@ -1,105 +1,105 @@
-# Clawdbot Integration
-
-AgentVibes TTS integration for Clawdbot - the AI assistant framework.
-
-## What is Clawdbot?
-
-Clawdbot is a powerful AI assistant framework that connects Claude AI to messaging platforms (WhatsApp, Telegram, Discord, etc.) and provides extensible skills.
-
-**Website**: https://clawd.bot
-**GitHub**: https://github.com/clawdbot/clawdbot
-
-## AgentVibes + Clawdbot
-
-This integration provides Clawdbot with professional TTS capabilities using AgentVibes:
-
-- **Free & Offline**: No API costs, works without internet
-- **50+ Voices**: Professional AI voices in 30+ languages
-- **Remote SSH Support**: Audio tunnels from server to local machine
-- **Zero Setup**: Automatic when AgentVibes is installed
-
-## Installation
-
-```bash
-npx agentvibes install
-```
-
-That's it! AgentVibes is ready to use.
-
-## Usage from Clawdbot
-
-Once installed, use AgentVibes for TTS:
-
-```bash
-npx agentvibes speak "Hello from Clawdbot"
-```
-
-## Features
-
-- **Automatic Detection**: Clawdbot automatically uses AgentVibes when available
-- **Voice Selection**: 50+ voices across 30+ languages
-- **SSH Tunneling**: Audio plays on local machine from remote servers
-- **Zero Config**: Works out of the box once AgentVibes is installed
-
-## Remote SSH Audio
-
-When running Clawdbot on a remote server, set up PulseAudio tunneling so audio plays on your local machine:
-
-**Quick Setup:**
-
-1. **Remote server** (`~/.bashrc`):
-```bash
-export PULSE_SERVER=tcp:localhost:14713
-```
-
-2. **Local machine** (`~/.ssh/config`):
-```
-Host your-server
-    RemoteForward 14713 localhost:14713
-```
-
-3. **Connect and test**:
-```bash
-ssh your-server
-agentvibes speak "Testing remote audio"
-```
-
-**Full guide**: https://github.com/paulpreibisch/AgentVibes/blob/master/docs/remote-audio-setup.md
-
-## Documentation
-
-- **Skill Documentation**: [skill/SKILL.md](skill/SKILL.md)
-- **AgentVibes Docs**: https://agentvibes.org
-- **Clawdbot Docs**: https://docs.clawd.bot
-
-## Architecture
-
-```
-Clawdbot Request
-    ↓
-AgentVibes Skill (this)
-    ↓
-AgentVibes CLI (agentvibes speak)
-    ↓
-Piper TTS Engine
-    ↓
-PulseAudio → SSH Tunnel → Local Speakers
-```
-
-## Development
-
-The skill is automatically packaged with AgentVibes npm releases. To update:
-
-1. Edit `skill/SKILL.md`
-2. Test with Clawdbot
-3. Commit to AgentVibes repo
-4. Publish new AgentVibes version
-
-## Support
-
-- **AgentVibes Issues**: https://github.com/paulpreibisch/AgentVibes/issues
-- **Clawdbot Issues**: https://github.com/clawdbot/clawdbot/issues
-
-## License
-
-Apache 2.0 - Same as AgentVibes
+# Clawdbot Integration
+
+AgentVibes TTS integration for Clawdbot - the AI assistant framework.
+
+## What is Clawdbot?
+
+Clawdbot is a powerful AI assistant framework that connects Claude AI to messaging platforms (WhatsApp, Telegram, Discord, etc.) and provides extensible skills.
+
+**Website**: https://clawd.bot
+**GitHub**: https://github.com/clawdbot/clawdbot
+
+## AgentVibes + Clawdbot
+
+This integration provides Clawdbot with professional TTS capabilities using AgentVibes:
+
+- **Free & Offline**: No API costs, works without internet
+- **50+ Voices**: Professional AI voices in 30+ languages
+- **Remote SSH Support**: Audio tunnels from server to local machine
+- **Zero Setup**: Automatic when AgentVibes is installed
+
+## Installation
+
+```bash
+npx agentvibes install
+```
+
+That's it! AgentVibes is ready to use.
+
+## Usage from Clawdbot
+
+Once installed, use AgentVibes for TTS:
+
+```bash
+npx agentvibes speak "Hello from Clawdbot"
+```
+
+## Features
+
+- **Automatic Detection**: Clawdbot automatically uses AgentVibes when available
+- **Voice Selection**: 50+ voices across 30+ languages
+- **SSH Tunneling**: Audio plays on local machine from remote servers
+- **Zero Config**: Works out of the box once AgentVibes is installed
+
+## Remote SSH Audio
+
+When running Clawdbot on a remote server, set up PulseAudio tunneling so audio plays on your local machine:
+
+**Quick Setup:**
+
+1. **Remote server** (`~/.bashrc`):
+```bash
+export PULSE_SERVER=tcp:localhost:14713
+```
+
+2. **Local machine** (`~/.ssh/config`):
+```
+Host your-server
+    RemoteForward 14713 localhost:14713
+```
+
+3. **Connect and test**:
+```bash
+ssh your-server
+agentvibes speak "Testing remote audio"
+```
+
+**Full guide**: https://github.com/paulpreibisch/AgentVibes/blob/master/docs/remote-audio-setup.md
+
+## Documentation
+
+- **Skill Documentation**: [skill/SKILL.md](skill/SKILL.md)
+- **AgentVibes Docs**: https://agentvibes.org
+- **Clawdbot Docs**: https://docs.clawd.bot
+
+## Architecture
+
+```
+Clawdbot Request
+    ↓
+AgentVibes Skill (this)
+    ↓
+AgentVibes CLI (agentvibes speak)
+    ↓
+Piper TTS Engine
+    ↓
+PulseAudio → SSH Tunnel → Local Speakers
+```
+
+## Development
+
+The skill is automatically packaged with AgentVibes npm releases. To update:
+
+1. Edit `skill/SKILL.md`
+2. Test with Clawdbot
+3. Commit to AgentVibes repo
+4. Publish new AgentVibes version
+
+## Support
+
+- **AgentVibes Issues**: https://github.com/paulpreibisch/AgentVibes/issues
+- **Clawdbot Issues**: https://github.com/clawdbot/clawdbot/issues
+
+## License
+
+Apache 2.0 - Same as AgentVibes