agentvibes 5.1.4 → 5.2.1

package/.claude/hooks/piper-multispeaker-registry.sh

@@ -1,171 +1,171 @@
-#!/usr/bin/env bash
-#
-# File: .claude/hooks/piper-multispeaker-registry.sh
-#
-# AgentVibes - Finally, your AI Agents can Talk Back! Text-to-Speech WITH personality for AI Assistants!
-# 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.
-#
-# DISCLAIMER: This software is provided "AS IS", WITHOUT WARRANTY OF ANY KIND,
-# express or implied, including but not limited to the warranties of
-# merchantability, fitness for a particular purpose and noninfringement.
-# In no event shall the authors or copyright holders be liable for any claim,
-# damages or other liability, whether in an action of contract, tort or
-# otherwise, arising from, out of or in connection with the software or the
-# use or other dealings in the software.
-#
-# ---
-#
-# @fileoverview Multi-Speaker Voice Registry - Maps speaker names to ONNX models and speaker IDs
-# @context Enables individual speaker selection from multi-speaker Piper models (e.g., 16Speakers)
-# @architecture Static registry mapping speaker names to model files and numeric speaker IDs
-# @dependencies piper-voice-manager.sh (voice storage), play-tts-piper.sh (TTS with speaker ID)
-# @entrypoints Sourced by voice-manager.sh for multi-speaker voice switching
-# @patterns Registry pattern, speaker ID mapping, model-to-speaker association
-# @related voice-manager.sh, play-tts-piper.sh, 16Speakers.onnx.json (speaker_id_map)
-#
-
-# Bash 3.2 compatible lowercase function (macOS ships with bash 3.2)
-# ${var,,} syntax requires bash 4.0+
-_to_lower() {
-  echo "$1" | tr '[:upper:]' '[:lower:]'
-}
-
-# Registry of multi-speaker models and their speaker names
-# Format: "SpeakerName:model_file:speaker_id:description"
-#
-# 16Speakers Model (12 US + 4 UK voices):
-# Source: LibriVox Public Domain recordings
-# Model: 16Speakers.onnx (77MB)
-#
-MULTISPEAKER_VOICES=(
-  # US English Speakers (0-11)
-  "Cori_Samuel:16Speakers:0:US English Female"
-  "Kara_Shallenberg:16Speakers:1:US English Female"
-  "Kristin_Hughes:16Speakers:2:US English Female"
-  "Maria_Kasper:16Speakers:3:US English Female"
-  "Mike_Pelton:16Speakers:4:US English Male"
-  "Mark_Nelson:16Speakers:5:US English Male"
-  "Michael_Scherer:16Speakers:6:US English Male"
-  "James_K_White:16Speakers:7:US English Male"
-  "Rose_Ibex:16Speakers:8:US English Female"
-  "progressingamerica:16Speakers:9:US English Male"
-  "Steve_C:16Speakers:10:US English Male"
-  "Owlivia:16Speakers:11:US English Female"
-
-  # UK English Speakers (12-15)
-  "Paul_Hampton:16Speakers:12:UK English Male"
-  "Jennifer_Dorr:16Speakers:13:UK English Female"
-  "Emily_Cripps:16Speakers:14:UK English Female"
-  "Martin_Clifton:16Speakers:15:UK English Male"
-)
-
-# @function get_multispeaker_info
-# @intent Get model and speaker ID for a speaker name
-# @why Enables users to select individual speakers from multi-speaker models by name
-# @param $1 {string} speaker_name - Speaker name (e.g., "Cori_Samuel", "Rose_Ibex")
-# @returns Echoes "model:speaker_id" (e.g., "16Speakers:0") to stdout
-# @exitcode 0=speaker found, 1=speaker not found
-# @sideeffects None (read-only lookup)
-# @edgecases Case-insensitive matching
-# @calledby voice-manager.sh switch command
-# @calls None (pure bash array iteration)
-get_multispeaker_info() {
-  local speaker_name="$1"
-  for entry in "${MULTISPEAKER_VOICES[@]}"; do
-    name="${entry%%:*}"
-    rest="${entry#*:}"
-    model="${rest%%:*}"
-    rest="${rest#*:}"
-    speaker_id="${rest%%:*}"
-
-    if [[ "$(_to_lower "$name")" == "$(_to_lower "$speaker_name")" ]]; then
-      echo "$model:$speaker_id"
-      return 0
-    fi
-  done
-  return 1
-}
-
-# @function list_multispeaker_voices
-# @intent Display all available multi-speaker voices with descriptions
-# @why Help users discover individual speakers within multi-speaker models
-# @param None
-# @returns None
-# @exitcode Always 0
-# @sideeffects Writes formatted list to stdout
-# @edgecases None
-# @calledby voice-manager.sh list command, /agent-vibes:list
-# @calls None (pure bash array iteration)
-list_multispeaker_voices() {
-  echo "🎭 Multi-Speaker Voices (16Speakers Model):"
-  echo "━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━"
-
-  local current_model=""
-  for entry in "${MULTISPEAKER_VOICES[@]}"; do
-    name="${entry%%:*}"
-    rest="${entry#*:}"
-    model="${rest%%:*}"
-    rest="${rest#*:}"
-    speaker_id="${rest%%:*}"
-    description="${rest#*:}"
-
-    # Print section header when model changes
-    if [[ "$model" != "$current_model" ]]; then
-      if [[ -n "$current_model" ]]; then
-        echo ""
-      fi
-      echo "  Model: $model.onnx"
-      current_model="$model"
-    fi
-
-    echo "    • $name (ID: $speaker_id) - $description"
-  done
-
-  echo "━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━"
-  echo ""
-  echo "Usage: /agent-vibes:switch Cori_Samuel"
-  echo "       /agent-vibes:switch Rose_Ibex"
-}
-
-# @function get_multispeaker_description
-# @intent Get description for a speaker name
-# @why Provide user-friendly info about speaker characteristics
-# @param $1 {string} speaker_name - Speaker name
-# @returns Echoes description (e.g., "US English Female") to stdout
-# @exitcode 0=speaker found, 1=speaker not found
-# @sideeffects None (read-only lookup)
-# @edgecases Case-insensitive matching
-# @calledby voice-manager.sh switch command (for confirmation message)
-# @calls None (pure bash array iteration)
-get_multispeaker_description() {
-  local speaker_name="$1"
-  for entry in "${MULTISPEAKER_VOICES[@]}"; do
-    name="${entry%%:*}"
-    rest="${entry#*:}"
-    rest="${rest#*:}"
-    rest="${rest#*:}"
-    description="${rest}"
-
-    if [[ "$(_to_lower "$name")" == "$(_to_lower "$speaker_name")" ]]; then
-      echo "$description"
-      return 0
-    fi
-  done
-  return 1
-}
+#!/usr/bin/env bash
+#
+# File: .claude/hooks/piper-multispeaker-registry.sh
+#
+# AgentVibes - Finally, your AI Agents can Talk Back! Text-to-Speech WITH personality for AI Assistants!
+# 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.
+#
+# DISCLAIMER: This software is provided "AS IS", WITHOUT WARRANTY OF ANY KIND,
+# express or implied, including but not limited to the warranties of
+# merchantability, fitness for a particular purpose and noninfringement.
+# In no event shall the authors or copyright holders be liable for any claim,
+# damages or other liability, whether in an action of contract, tort or
+# otherwise, arising from, out of or in connection with the software or the
+# use or other dealings in the software.
+#
+# ---
+#
+# @fileoverview Multi-Speaker Voice Registry - Maps speaker names to ONNX models and speaker IDs
+# @context Enables individual speaker selection from multi-speaker Piper models (e.g., 16Speakers)
+# @architecture Static registry mapping speaker names to model files and numeric speaker IDs
+# @dependencies piper-voice-manager.sh (voice storage), play-tts-piper.sh (TTS with speaker ID)
+# @entrypoints Sourced by voice-manager.sh for multi-speaker voice switching
+# @patterns Registry pattern, speaker ID mapping, model-to-speaker association
+# @related voice-manager.sh, play-tts-piper.sh, 16Speakers.onnx.json (speaker_id_map)
+#
+
+# Bash 3.2 compatible lowercase function (macOS ships with bash 3.2)
+# ${var,,} syntax requires bash 4.0+
+_to_lower() {
+  echo "$1" | tr '[:upper:]' '[:lower:]'
+}
+
+# Registry of multi-speaker models and their speaker names
+# Format: "SpeakerName:model_file:speaker_id:description"
+#
+# 16Speakers Model (12 US + 4 UK voices):
+# Source: LibriVox Public Domain recordings
+# Model: 16Speakers.onnx (77MB)
+#
+MULTISPEAKER_VOICES=(
+  # US English Speakers (0-11)
+  "Cori_Samuel:16Speakers:0:US English Female"
+  "Kara_Shallenberg:16Speakers:1:US English Female"
+  "Kristin_Hughes:16Speakers:2:US English Female"
+  "Maria_Kasper:16Speakers:3:US English Female"
+  "Mike_Pelton:16Speakers:4:US English Male"
+  "Mark_Nelson:16Speakers:5:US English Male"
+  "Michael_Scherer:16Speakers:6:US English Male"
+  "James_K_White:16Speakers:7:US English Male"
+  "Rose_Ibex:16Speakers:8:US English Female"
+  "progressingamerica:16Speakers:9:US English Male"
+  "Steve_C:16Speakers:10:US English Male"
+  "Owlivia:16Speakers:11:US English Female"
+
+  # UK English Speakers (12-15)
+  "Paul_Hampton:16Speakers:12:UK English Male"
+  "Jennifer_Dorr:16Speakers:13:UK English Female"
+  "Emily_Cripps:16Speakers:14:UK English Female"
+  "Martin_Clifton:16Speakers:15:UK English Male"
+)
+
+# @function get_multispeaker_info
+# @intent Get model and speaker ID for a speaker name
+# @why Enables users to select individual speakers from multi-speaker models by name
+# @param $1 {string} speaker_name - Speaker name (e.g., "Cori_Samuel", "Rose_Ibex")
+# @returns Echoes "model:speaker_id" (e.g., "16Speakers:0") to stdout
+# @exitcode 0=speaker found, 1=speaker not found
+# @sideeffects None (read-only lookup)
+# @edgecases Case-insensitive matching
+# @calledby voice-manager.sh switch command
+# @calls None (pure bash array iteration)
+get_multispeaker_info() {
+  local speaker_name="$1"
+  for entry in "${MULTISPEAKER_VOICES[@]}"; do
+    name="${entry%%:*}"
+    rest="${entry#*:}"
+    model="${rest%%:*}"
+    rest="${rest#*:}"
+    speaker_id="${rest%%:*}"
+
+    if [[ "$(_to_lower "$name")" == "$(_to_lower "$speaker_name")" ]]; then
+      echo "$model:$speaker_id"
+      return 0
+    fi
+  done
+  return 1
+}
+
+# @function list_multispeaker_voices
+# @intent Display all available multi-speaker voices with descriptions
+# @why Help users discover individual speakers within multi-speaker models
+# @param None
+# @returns None
+# @exitcode Always 0
+# @sideeffects Writes formatted list to stdout
+# @edgecases None
+# @calledby voice-manager.sh list command, /agent-vibes:list
+# @calls None (pure bash array iteration)
+list_multispeaker_voices() {
+  echo "🎭 Multi-Speaker Voices (16Speakers Model):"
+  echo "━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━"
+
+  local current_model=""
+  for entry in "${MULTISPEAKER_VOICES[@]}"; do
+    name="${entry%%:*}"
+    rest="${entry#*:}"
+    model="${rest%%:*}"
+    rest="${rest#*:}"
+    speaker_id="${rest%%:*}"
+    description="${rest#*:}"
+
+    # Print section header when model changes
+    if [[ "$model" != "$current_model" ]]; then
+      if [[ -n "$current_model" ]]; then
+        echo ""
+      fi
+      echo "  Model: $model.onnx"
+      current_model="$model"
+    fi
+
+    echo "    • $name (ID: $speaker_id) - $description"
+  done
+
+  echo "━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━"
+  echo ""
+  echo "Usage: /agent-vibes:switch Cori_Samuel"
+  echo "       /agent-vibes:switch Rose_Ibex"
+}
+
+# @function get_multispeaker_description
+# @intent Get description for a speaker name
+# @why Provide user-friendly info about speaker characteristics
+# @param $1 {string} speaker_name - Speaker name
+# @returns Echoes description (e.g., "US English Female") to stdout
+# @exitcode 0=speaker found, 1=speaker not found
+# @sideeffects None (read-only lookup)
+# @edgecases Case-insensitive matching
+# @calledby voice-manager.sh switch command (for confirmation message)
+# @calls None (pure bash array iteration)
+get_multispeaker_description() {
+  local speaker_name="$1"
+  for entry in "${MULTISPEAKER_VOICES[@]}"; do
+    name="${entry%%:*}"
+    rest="${entry#*:}"
+    rest="${rest#*:}"
+    rest="${rest#*:}"
+    description="${rest}"
+
+    if [[ "$(_to_lower "$name")" == "$(_to_lower "$speaker_name")" ]]; then
+      echo "$description"
+      return 0
+    fi
+  done
+  return 1
+}

package/.claude/hooks/piper-voice-manager.sh

@@ -265,10 +265,135 @@ download_voice() {
     return 1
   fi
 
+  # Patch LibriTTS speaker names if this is a libritts model
+  if [[ "$voice_name" == *libritts* ]]; then
+    patch_libritts_speaker_names "$voice_dir" "$voice_name"
+  fi
+
   echo "✅ Voice downloaded successfully: $voice_name"
   echo "   Location: $voice_dir/${voice_name}.onnx"
 }
 
+# @function patch_libritts_speaker_names
+# @intent Replace raw corpus IDs (p3922, p8699) with friendly names (Anna, Bella) in LibriTTS .onnx.json
+# @why Users see cryptic "p100Bell" names instead of friendly names without this patch
+# @param $1 {string} voice_dir - Directory containing the .onnx.json file
+# @param $2 {string} voice_name - Voice model name (e.g., en_US-libritts-high)
+# @returns None
+# @exitcode Always 0 (non-fatal)
+# @sideeffects Rewrites speaker_id_map in .onnx.json with friendly names from voice-assignments.json
+# @calledby download_voice (for libritts models), piper-download-voices.sh (post-download)
+# @calls python3/node (for JSON manipulation)
+patch_libritts_speaker_names() {
+  local voice_dir="$1"
+  local voice_name="$2"
+  local json_file="$voice_dir/${voice_name}.onnx.json"
+
+  if [[ ! -f "$json_file" ]]; then
+    return 0
+  fi
+
+  # Find voice-assignments.json relative to this script (SCRIPT_DIR/../.. = project root)
+  local script_dir
+  script_dir="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+  local project_root
+  project_root="$(cd "$script_dir/../.." 2>/dev/null && pwd)"
+  local catalog="$project_root/voice-assignments.json"
+
+  # Also check npm global install location
+  if [[ ! -f "$catalog" ]]; then
+    # Try npm root
+    local npm_root
+    npm_root="$(npm root -g 2>/dev/null)/agentvibes" || true
+    if [[ -f "$npm_root/voice-assignments.json" ]]; then
+      catalog="$npm_root/voice-assignments.json"
+    fi
+  fi
+
+  if [[ ! -f "$catalog" ]]; then
+    return 0
+  fi
+
+  # Check if already patched (first key doesn't start with 'p' + digits)
+  if command -v python3 &>/dev/null; then
+    python3 -c "
+import json, sys
+
+json_path = sys.argv[1]
+catalog_path = sys.argv[2]
+
+with open(json_path, 'r') as f:
+    data = json.load(f)
+
+sid_map = data.get('speaker_id_map', {})
+if not sid_map or data.get('num_speakers', 0) <= 1:
+    sys.exit(0)
+
+# Check if already patched
+import re
+first_key = next(iter(sid_map))
+if not re.match(r'^p\d+$', first_key):
+    sys.exit(0)
+
+# Load catalog
+with open(catalog_path, 'r') as f:
+    catalog = json.load(f)
+
+speakers = catalog.get('libritts_speakers', {})
+
+# Build reverse map: index -> p-name
+index_to_p = {v: k for k, v in sid_map.items()}
+
+# Rebuild with friendly names
+new_map = {}
+for idx, pname in index_to_p.items():
+    friendly = speakers.get(str(idx), {}).get('voice_name')
+    new_map[friendly if friendly else pname] = idx
+
+data['speaker_id_map'] = new_map
+
+with open(json_path, 'w') as f:
+    json.dump(data, f, indent=2)
+
+print('   Patched LibriTTS speaker names to friendly names')
+" "$json_file" "$catalog" 2>/dev/null || true
+  elif command -v node &>/dev/null; then
+    node -e "
+const fs = require('fs');
+const path = require('path');
+
+const jsonPath = process.argv[1];
+const catalogPath = process.argv[2];
+
+const data = JSON.parse(fs.readFileSync(jsonPath, 'utf8'));
+const sidMap = data.speaker_id_map || {};
+if (!Object.keys(sidMap).length || (data.num_speakers || 0) <= 1) process.exit(0);
+
+// Check if already patched
+const firstKey = Object.keys(sidMap)[0];
+if (!/^p\d+$/.test(firstKey)) process.exit(0);
+
+const catalog = JSON.parse(fs.readFileSync(catalogPath, 'utf8'));
+const speakers = catalog.libritts_speakers || {};
+
+// Build reverse map
+const indexToP = {};
+for (const [pname, idx] of Object.entries(sidMap)) indexToP[idx] = pname;
+
+// Rebuild with friendly names
+const newMap = {};
+for (const [idx, pname] of Object.entries(indexToP)) {
+  const friendly = speakers[String(idx)]?.voice_name;
+  newMap[friendly || pname] = parseInt(idx, 10);
+}
+
+data.speaker_id_map = newMap;
+fs.writeFileSync(jsonPath, JSON.stringify(data, null, 2), 'utf8');
+console.log('   Patched LibriTTS speaker names to friendly names');
+" "$json_file" "$catalog" 2>/dev/null || true
+  fi
+}
+
 # @function list_downloaded_voices
 # @intent Display all locally cached voice models with file sizes
 # @why Help users see what voices they have available and storage usage