* fix: support no_transcript option for demo

* fix: fix manifest tye

Avatar extensions (anam, generic, heygen):
- Replace 500ms debounce timer with tts_audio_end event handling
- Add send_voice_end() method triggered by tts_audio_end (reason=1)
- Remove dead code from heygen (unused queues, is_speaking timeout)
- Fix graph routing: tts_audio_end now routes to avatar (was only thymia)

Deepgram WebSocket ASR:
- Add silence sender for EOT detection when mic muted
- Add auto-reconnect with exponential backoff
- Add keep-alive silence to prevent WebSocket timeout

Playground:
- Add URL param support: ?graph=name pre-selects graph on load

Co-authored-by: Ubuntu <[email protected]>
Co-authored-by: Ethan Zhang <[email protected]>

…1883)

* feat: add hellos graph and share code between thymia analysis modes

- Add flux_hellos_gpt_5_1_cartesia_anam graph for hellos-only analysis
- Refactor thymia_analyzer to share code between hellos_only and full modes
- Set min_speech_duration to 10s for faster demo turnaround

* fix: sync rebuild_property.py with 10s min_speech_duration

---------

Co-authored-by: Ubuntu <[email protected]>
Co-authored-by: Ethan Zhang <[email protected]>

* init voice-assistant-with-PowerMem

* refactor: replace Memu memory store with PowerMem integration

- Updated memory store implementation to use PowerMemSdkMemoryStore.
- Simplified initialization by removing self-hosting options.
- Enhanced logging for memory operations and error handling.
- Updated requirements to include powermem instead of memu-py.

* refactor: update voice assistant configuration for Azure integration

- Replaced Deepgram ASR and ElevenLabs TTS with Azure ASR and TTS services.
- Updated property keys and parameters to align with Azure's requirements.
- Removed unused fields such as agent_name and user_name from configuration.
- Simplified memorization logic by eliminating user_name parameter from memory functions.

* refactor: update voice assistant to use PowerMem and enhance documentation

- Added .gitignore to exclude tenapp/data/ directory.
- Updated Dockerfile to reference the new voice assistant with PowerMem.
- Renamed memory-related attributes in extension.py for clarity.
- Revised README files to reflect the transition from MemU to PowerMem, including updated features and installation instructions.

* feat: enhance voice assistant with personalized greetings and memory integration

- Updated greeting message to reflect memory capabilities.
- Implemented personalized greeting generation based on user memories.
- Adjusted memorization logic to occur every five rounds instead of two.
- Added error handling and logging for personalized greeting generation.

* feat: implement personalized greeting handling in voice assistant

- Added state management for greeting generation in MainControlExtension.
- Enhanced logic to generate and log personalized greetings based on user memories.
- Updated response handling to manage final and non-final LLM responses during greeting generation.
- Ensured default greetings are used when personalized greetings are not available.

* refactor: simplify memorization logic in voice assistant

- Updated comment to clarify that memorization occurs every five rounds if enabled.
- Removed redundant condition from the comment for better readability.

* fix: improve personalized greeting handling in MainControlExtension

- Ensured personalized greetings are set only if the config is available.
- Streamlined the logic for sending TTS and transcript messages based on greeting availability.
- Enhanced logging for cases when no personalized greeting is generated.

* refactor: streamline memory retrieval methods in MainControlExtension

- Removed user_id parameter from memory retrieval functions for simplification.
- Updated calls to memory retrieval methods to align with the new signature.
- Enhanced code clarity by reducing unnecessary parameters in related memory functions.

* chore: update manifest-lock and property files for voice assistant

- Updated versions and hashes for several dependencies in manifest-lock.json to the latest releases.
- Added language parameter to Azure STT configuration in property.json for improved multilingual support.
- Enhanced TTS configuration by specifying the default synthesis language in property.json.

* refactor: comment out memory loading in MainControlExtension

- Commented out the call to _load_memory_to_context() for potential debugging or future implementation considerations.
- Maintained existing functionality while allowing for easier testing of other components.

* refactor: update memory handling in MainControlExtension and PowerMemSdkMemoryStore

- Removed the MemoryStore base class and its associated methods, simplifying the PowerMemSdkMemoryStore implementation.
- Updated MainControlExtension to directly use PowerMemSdkMemoryStore for memory operations.
- Enhanced the context message format for related memory retrieval to improve user interaction.
- Streamlined memory retrieval logic by optimizing response handling and removing unnecessary parameters.

* refactor: enhance memory handling in MainControlExtension and PowerMemSdkMemoryStore

- Introduced a new method to memorize conversations in MainControlExtension, improving memory integration.
- Renamed memory store methods for clarity, changing 'memorize' to 'add' and 'retrieve_related_clustered_categories' to 'search'.
- Removed unused memory summary parsing methods to streamline the codebase and improve maintainability.

* refactor: remove unused memory methods in MainControlExtension and update logging in PowerMemSdkMemoryStore

- Commented out the memory loading method in MainControlExtension to streamline functionality.
- Removed the call to _memorize_conversation() to enhance clarity and maintainability.
- Updated logging in PowerMemSdkMemoryStore to reflect method name change from 'memorize' to 'add' for better understanding.

* feat: enhance voice assistant with PowerMem memory management

- Added comprehensive PowerMem configuration options to the main README, detailing database, LLM, and embedding settings.
- Updated the extension README to reflect the integration of PowerMem, highlighting features like conversation memory, semantic search, and personalized greetings.
- Included installation instructions and environment configuration guidelines for seamless setup.
- Enhanced memory management workflow and error handling to improve user experience and functionality.

* fix: update memorization frequency in voice assistant

- Changed the memorization logic to save conversation to PowerMem every 2 rounds instead of 5, enhancing memory retention.
- Updated README documentation to reflect the new memorization frequency and its implications for memory management.

* feat: integrate templates for context messages and personalized greetings in voice assistant

- Replaced hardcoded context message and greeting prompt with template constants for improved maintainability and readability.
- Introduced CONTEXT_MESSAGE_WITH_MEMORY_TEMPLATE and PERSONALIZED_GREETING_TEMPLATE to streamline message formatting and enhance code clarity.

* fix: refine memorization logic in MainControlExtension

- Updated the memorization condition to check the difference between the current turn ID and the last memory update turn ID, ensuring conversations are memorized every two rounds when enabled.
- Added a new attribute to track the last memory update turn ID for improved memory management.

* feat: enhance memory management in voice assistant

- Introduced configurable memory saving rules, allowing the assistant to save conversations based on a specified number of turns or after a period of inactivity.
- Updated the configuration to include `memory_save_interval_turns` and `memory_idle_timeout_seconds` for improved flexibility in memory management.
- Enhanced the README documentation to reflect these new features and their usage, ensuring users can easily configure memory settings.

* refactor: improve memory store abstraction in voice assistant

- Introduced a new MemoryStore base class to enhance the memory management structure.
- Updated MainControlExtension to utilize MemoryStore instead of directly using PowerMemSdkMemoryStore.
- Renamed parameters in memory search methods for consistency and clarity, changing 'category_query' to 'query'.
- Revised README documentation to reflect changes in memory retrieval parameters and class structure.

* feat: update personalized greeting template in voice assistant

- Enhanced the PERSONALIZED_GREETING_TEMPLATE to include a condition for responding in the most commonly used language of the user's location/region if such information is available in the memory summary.
- This change aims to improve user engagement by making greetings more contextually relevant and personalized.

* feat: refine personalized greeting template in voice assistant

- Updated the PERSONALIZED_GREETING_TEMPLATE to emphasize the assistant's friendly and helpful nature, enhancing user interaction.
- This change aims to create a more engaging and contextually relevant greeting experience for users.

* fix: correct typographical errors in prompt templates for voice assistant

- Updated the CONTEXT_MESSAGE_WITH_MEMORY_TEMPLATE and PERSONALIZED_GREETING_TEMPLATE to fix typographical inconsistencies, ensuring proper use of apostrophes.
- This change enhances the clarity and professionalism of the prompt strings used in the extension.

* feat: add user memory management to voice assistant

- Introduced `enable_user_memory` configuration option to control user-specific memory functionality.
- Updated `MainControlExtension` to initialize the appropriate memory store based on the new configuration.
- Added `PowerMemSdkUserMemoryStore` class for handling user-specific memory operations.
- Revised documentation to include the new configuration parameter and its implications for memory management.

* refactor: update type hinting for conversation parameter in memory store classes

- Changed type hint for the `conversation` parameter from `List[dict]` to `list[dict]` in the `add` method across all memory store classes for consistency with Python's type hinting conventions.
- This change improves code clarity and aligns with modern Python practices.

* docs: enhance README documentation for voice assistant with PowerMem

- Updated README files to include links to the PowerMem repository, improving accessibility to relevant resources.
- Clarified the description of PowerMem's memory management capabilities in the context of the voice assistant, enhancing user understanding of its features.

* feat: add user memory configuration options to manifest

- Introduced new configuration options in the manifest.json for user memory management, including `enable_user_memory`, `memory_save_interval_turns`, and `memory_idle_timeout_seconds`.
- These additions enhance the flexibility and control over user-specific memory functionalities in the voice assistant.

* feat: implement user profile retrieval in memory store

- Added a new abstract method `get_user_profile` to the MemoryStore class for fetching user profiles.
- Implemented the `get_user_profile` method in the PowerMemSdkMemoryStore class to retrieve and format user profile information.
- Updated the MainControlExtension to utilize the new method for accessing user memory summaries, enhancing the assistant's ability to provide personalized interactions.

* fix: correct user memory initialization logic in MainControlExtension

- Updated the condition for initializing the memory store in MainControlExtension to ensure that the PowerMemSdkMemoryStore is used when user memory is disabled, aligning with the intended configuration behavior.
- This change improves the accuracy of memory management based on user settings.

* chore: remove .gitignore file for voice assistant with PowerMem

- Deleted the .gitignore file from the voice assistant example directory, which previously excluded the 'tenapp/data/' directory. This change simplifies the project structure by removing unnecessary files.

* fix: update language configuration in voice assistant property file

- Modified the language parameter in the property.json file to specify only "en-US", removing "zh-CN". This change streamlines language support for the voice assistant, ensuring it operates with a single language setting.

* fix: update dependencies and configuration in voice assistant files

- Downgraded versions and updated hashes for several dependencies in manifest-lock.json to ensure compatibility.
- Changed the STT addon from "azure_asr_python" to "deepgram_asr_python" and updated its parameters in property.json for improved speech recognition.
- Switched the TTS addon from "azure_tts_python" to "elevenlabs_tts2_python" and modified its configuration for better performance and output quality.

* fix: enhance memory store initialization error handling in MainControlExtension

- Added error handling during the initialization of the PowerMem memory store to log failures and allow the extension to continue without memory functionality if initialization fails.
- Improved logging to provide detailed information about the initialization process and any exceptions encountered, enhancing debugging capabilities.

* fix: update logging in PowerMemSdkUserMemoryStore for clarity

- Changed log messages in the PowerMemSdkUserMemoryStore class to reflect the correct class name, enhancing clarity in log outputs.
- This update ensures that logs accurately represent the source of memory operations, aiding in debugging and monitoring.

* fix: improve logging and error handling in user profile retrieval methods

- Enhanced logging in the `get_user_profile` and `get_user_profile_by_query` methods of both `PowerMemSdkMemoryStore` and `PowerMemSdkUserMemoryStore` classes to provide detailed information about the retrieval process and results.
- Added error handling to log exceptions and tracebacks, improving the robustness of user profile retrieval and aiding in debugging efforts.

* fix: simplify user profile retrieval in MainControlExtension

- Removed the query parameter from the `get_user_profile` method call in MainControlExtension, allowing for a more straightforward retrieval of user memory summaries.
- This change enhances the clarity and efficiency of the memory retrieval process.

* fix: refine user profile retrieval parameters in MainControlExtension

- Updated the `get_user_profile` method call in MainControlExtension to include user_id and agent_id parameters, enhancing the specificity of user memory retrieval.
- This change improves the accuracy of the memory summary fetched for the user, ensuring personalized interactions.

* fix: remove agent_id parameter from user profile retrieval in PowerMemSdkUserMemoryStore

- Eliminated the agent_id parameter from the profile retrieval call in the PowerMemSdkUserMemoryStore class, streamlining the user profile fetching process.
- This change enhances the clarity and efficiency of user memory interactions, aligning with recent updates to improve memory retrieval methods.

* fix: update configuration settings in voice assistant README

- Added timezone configuration for better localization support.
- Renamed database configuration parameters for clarity and consistency.
- Introduced OceanBase and PostgreSQL configuration settings for enhanced database integration.
- Added optional graph store configuration to support knowledge graph functionality.
- Improved overall documentation for better user guidance on setup and configuration.

* fix: enable user memory functionality in voice assistant configuration

- Updated configuration files to enable user memory and memorization features for the voice assistant.
- Adjusted property.json and config.py files to reflect the changes, enhancing user profile management capabilities.
- Improved README documentation to clarify the new settings and their implications for user interactions.

* fix: update Dockerfile and manifest versions for voice assistant

- Updated the Dockerfile to use the latest builder image version 0.7.12.
- Changed the version specification for the ten_ai_base dependency in manifest.json files from "=0.7.22" to "0.7" to allow for more flexible version compatibility.

* feat: add Docker support for voice assistant with PowerMem

- Introduced a Dockerfile for building the voice assistant application.
- Added a docker-compose.yml file to manage services, including the voice assistant and SeekDB database.
- Created a .gitignore file to exclude data directory from version control.

* feat: update configuration for voice assistant with PowerMem

- Revised README files to enhance clarity on required and optional environment variables for the voice assistant.
- Changed the default database provider from SQLite to OceanBase in configuration templates.
- Removed unnecessary comments and default settings to streamline the configuration process.

* feat: enhance README for voice assistant with Docker instructions

- Added a new section in the README for Docker setup, including commands to start the container and access the application.
- Improved clarity on the quick start process for users deploying the voice assistant.

* docs: expand README for voice assistant with detailed setup instructions

- Enhanced the README with a comprehensive table of contents for easier navigation.
- Added detailed sections on prerequisites, configuration, and verification steps for both Docker and local development setups.
- Improved clarity on required environment variables and provided examples for configuration.
- Included troubleshooting tips and common issues to assist users in resolving potential setup problems.

* feat: add Chinese documentation for voice assistant with PowerMem

- Created a new README_CN.md file to provide comprehensive setup instructions and information in Chinese.
- Updated the main README.md to include a link to the Chinese documentation for better accessibility for Chinese-speaking users.

* refactor: remove Chinese README for voice assistant with PowerMem

- Deleted the README_CN.md file as it is no longer needed.
- Updated the main README.md to remove the link to the Chinese documentation, streamlining the content for English-speaking users.

* docs: update README for voice assistant with PowerMem

- Modified the "Quick Start" section header in the README to improve clarity by replacing the hyphen with three dashes.
- Ensured consistency in formatting for better readability.

* docs: add configuration guide for voice assistant with PowerMem

- Introduced a new CONFIGURATION.md file detailing environment variables and configuration options for the voice assistant.
- Updated README.md to include references to the new configuration guide, enhancing user support for setup and alternative providers.

* chore: update configuration and documentation for voice assistant with PowerMem

- Enhanced .gitignore to include .env file and data directory for better environment management.
- Updated docker-compose.yml to reference the local .env file for environment variables.
- Revised README.md to clarify steps for configuring environment variables and starting services, including new instructions for SeekDB container setup.
- Adjusted Taskfile.yml to use the local .env file for dotenv configuration.

* feat: update Dockerfile and requirements for voice assistant with PowerMem

- Added a step in the Dockerfile to copy and install Python dependencies from requirements.txt.
- Updated requirements.txt to specify a minimum version for the powermem package, ensuring compatibility.

* docs: remove outdated environment configuration from README for voice assistant with PowerMem

- Eliminated the detailed environment configuration section from README.md to streamline documentation.
- Focused on providing a concise overview of configuration options, enhancing clarity for users.

* docs: simplify service startup instructions in README for voice assistant with PowerMem

- Reorganized the steps for starting services with Docker Compose to improve clarity.
- Renumbered the steps to create a more logical flow, enhancing user experience.

* feat: add .env.example file and update README for voice assistant with PowerMem

- Introduced a new .env.example file containing environment variable configurations for the voice assistant.
- Updated README.md to reflect changes in environment setup instructions, including a more streamlined process for copying the .env.example file.
- Renumbered steps in the README to improve logical flow and clarity for users.

* docs: update README to remove unnecessary environment variable URLs for voice assistant with PowerMem

- Removed LLM_BASE_URL and EMBEDDING_BASE_URL from the README.md as they are no longer needed.
- Streamlined the documentation to focus on essential configuration options, enhancing clarity for users.

* docs: update .env.example and README for voice assistant with PowerMem

- Added new options for LLM_PROVIDER and EMBEDDING_PROVIDER in both .env.example and README.md to include ollama, vllm, anthropic, deepseek, huggingface, and lmstudio.
- Ensured consistency between the environment variable documentation and the actual configuration options available.

* feat: enhance PowerMem integration in voice assistant example

- Updated .env.example to use localhost for OCEANBASE_HOST.
- Added PowerMem configuration options in property.json for vector store, LLM, and embedder.
- Modified MainControlConfig to include powermem_config.
- Updated memory store initialization to utilize PowerMem configuration.
- Improved logging for configuration loading in extension.py.

* chore: remove obsolete files and update Dockerfile for voice assistant with PowerMem

- Deleted .gitignore and docker-compose.yml as they are no longer needed.
- Updated Dockerfile to copy .env.example to .env and added seekdb as a build stage.
- Modified Taskfile.docker.yml to include a new task for running the seekdb server in production.
- Adjusted Taskfile.yml to reference the correct .env path.

* refactor: remove .env.example and streamline README for voice assistant with PowerMem

- Deleted obsolete .env.example file to simplify configuration.
- Updated README.md to consolidate PowerMem configuration details and quick start instructions.
- Enhanced clarity on environment variable setup and added new configuration options for memory management in the extension.
- Improved documentation on memory functionality and user greeting generation.

* docs: update README for voice assistant with PowerMem to include seekdb server setup

- Added instructions to start the seekdb server as the first step in the quick start guide.
- Renumbered subsequent steps for clarity and consistency.
- Enhanced overall documentation for better user guidance.

* docs: update README for voice assistant with PowerMem to include ROOT_PASSWORD environment variable

- Added ROOT_PASSWORD environment variable to the seekdb server setup instructions for clarity.
- Ensured consistency with previous updates to the README for better user guidance.

* docs: refine README for voice assistant with PowerMem to enhance clarity on environment variables

- Removed comments from the .env section for a cleaner presentation.
- Ensured consistency in formatting for LLM and embedding provider sections.
- Improved overall readability of the configuration instructions.

* docs: remove CONFIGURATION.md for voice assistant with PowerMem

- Deleted the outdated configuration guide to streamline documentation.
- Consolidated environment variable and configuration details into the README for improved clarity and accessibility.

---------

Co-authored-by: Ethan Zhang <[email protected]>

* chore: calculate coverage in debug mode

* Revert "chore: calculate coverage in debug mode"

This reverts commit f68a031.

* Reapply "chore: calculate coverage in debug mode"

This reverts commit bc27f2c.

* fix: remove container for test jobs

* fix: add C++ toolchain, go builder, sanitizer support, nodejs env

* fix: compiler version incompatible

* Revert "fix: compiler version incompatible"

This reverts commit 23f5c3f.

* fix: set PATH so that clang-21 is find earlier than older versions

* fix: uninstall older versions of clang

* Revert "fix: set PATH so that clang-21 is find earlier than older versions"

This reverts commit b66e56c.

* fix: uninstall old clang first then install clang21

* fix: supplement test jobs deps according to tools Dockerfile

* fix: install libasan5 for all jobs and remove unecessary deps

* chore: refine codes and output clang version before every test

* fix: go back to clang 18 to detect __lsan_do_leak_check

* Revert "fix: go back to clang 18 to detect __lsan_do_leak_check"

This reverts commit d247818.

* fix: upgrade libasan5 to libasan8 to match clang 21

* fix: install libclang-rt-21-dev

* Revert "fix: install libclang-rt-21-dev"

This reverts commit 39b96e0.

* fix: ten_enable_go_app_leak_check

* chore: uninstall useless dep libasan8

* chore: refine codes and uninstall unecessary dep clang-tool-21

* init voice-assistant-with-PowerMem

* refactor: replace Memu memory store with PowerMem integration

- Updated memory store implementation to use PowerMemSdkMemoryStore.
- Simplified initialization by removing self-hosting options.
- Enhanced logging for memory operations and error handling.
- Updated requirements to include powermem instead of memu-py.

* refactor: update voice assistant configuration for Azure integration

- Replaced Deepgram ASR and ElevenLabs TTS with Azure ASR and TTS services.
- Updated property keys and parameters to align with Azure's requirements.
- Removed unused fields such as agent_name and user_name from configuration.
- Simplified memorization logic by eliminating user_name parameter from memory functions.

* refactor: update voice assistant to use PowerMem and enhance documentation

- Added .gitignore to exclude tenapp/data/ directory.
- Updated Dockerfile to reference the new voice assistant with PowerMem.
- Renamed memory-related attributes in extension.py for clarity.
- Revised README files to reflect the transition from MemU to PowerMem, including updated features and installation instructions.

* feat: enhance voice assistant with personalized greetings and memory integration

- Updated greeting message to reflect memory capabilities.
- Implemented personalized greeting generation based on user memories.
- Adjusted memorization logic to occur every five rounds instead of two.
- Added error handling and logging for personalized greeting generation.

* feat: implement personalized greeting handling in voice assistant

- Added state management for greeting generation in MainControlExtension.
- Enhanced logic to generate and log personalized greetings based on user memories.
- Updated response handling to manage final and non-final LLM responses during greeting generation.
- Ensured default greetings are used when personalized greetings are not available.

* refactor: simplify memorization logic in voice assistant

- Updated comment to clarify that memorization occurs every five rounds if enabled.
- Removed redundant condition from the comment for better readability.

* fix: improve personalized greeting handling in MainControlExtension

- Ensured personalized greetings are set only if the config is available.
- Streamlined the logic for sending TTS and transcript messages based on greeting availability.
- Enhanced logging for cases when no personalized greeting is generated.

* refactor: streamline memory retrieval methods in MainControlExtension

- Removed user_id parameter from memory retrieval functions for simplification.
- Updated calls to memory retrieval methods to align with the new signature.
- Enhanced code clarity by reducing unnecessary parameters in related memory functions.

* chore: update manifest-lock and property files for voice assistant

- Updated versions and hashes for several dependencies in manifest-lock.json to the latest releases.
- Added language parameter to Azure STT configuration in property.json for improved multilingual support.
- Enhanced TTS configuration by specifying the default synthesis language in property.json.

* refactor: comment out memory loading in MainControlExtension

- Commented out the call to _load_memory_to_context() for potential debugging or future implementation considerations.
- Maintained existing functionality while allowing for easier testing of other components.

* refactor: update memory handling in MainControlExtension and PowerMemSdkMemoryStore

- Removed the MemoryStore base class and its associated methods, simplifying the PowerMemSdkMemoryStore implementation.
- Updated MainControlExtension to directly use PowerMemSdkMemoryStore for memory operations.
- Enhanced the context message format for related memory retrieval to improve user interaction.
- Streamlined memory retrieval logic by optimizing response handling and removing unnecessary parameters.

* refactor: enhance memory handling in MainControlExtension and PowerMemSdkMemoryStore

- Introduced a new method to memorize conversations in MainControlExtension, improving memory integration.
- Renamed memory store methods for clarity, changing 'memorize' to 'add' and 'retrieve_related_clustered_categories' to 'search'.
- Removed unused memory summary parsing methods to streamline the codebase and improve maintainability.

* refactor: remove unused memory methods in MainControlExtension and update logging in PowerMemSdkMemoryStore

- Commented out the memory loading method in MainControlExtension to streamline functionality.
- Removed the call to _memorize_conversation() to enhance clarity and maintainability.
- Updated logging in PowerMemSdkMemoryStore to reflect method name change from 'memorize' to 'add' for better understanding.

* feat: enhance voice assistant with PowerMem memory management

- Added comprehensive PowerMem configuration options to the main README, detailing database, LLM, and embedding settings.
- Updated the extension README to reflect the integration of PowerMem, highlighting features like conversation memory, semantic search, and personalized greetings.
- Included installation instructions and environment configuration guidelines for seamless setup.
- Enhanced memory management workflow and error handling to improve user experience and functionality.

* fix: update memorization frequency in voice assistant

- Changed the memorization logic to save conversation to PowerMem every 2 rounds instead of 5, enhancing memory retention.
- Updated README documentation to reflect the new memorization frequency and its implications for memory management.

* feat: integrate templates for context messages and personalized greetings in voice assistant

- Replaced hardcoded context message and greeting prompt with template constants for improved maintainability and readability.
- Introduced CONTEXT_MESSAGE_WITH_MEMORY_TEMPLATE and PERSONALIZED_GREETING_TEMPLATE to streamline message formatting and enhance code clarity.

* fix: refine memorization logic in MainControlExtension

- Updated the memorization condition to check the difference between the current turn ID and the last memory update turn ID, ensuring conversations are memorized every two rounds when enabled.
- Added a new attribute to track the last memory update turn ID for improved memory management.

* feat: enhance memory management in voice assistant

- Introduced configurable memory saving rules, allowing the assistant to save conversations based on a specified number of turns or after a period of inactivity.
- Updated the configuration to include `memory_save_interval_turns` and `memory_idle_timeout_seconds` for improved flexibility in memory management.
- Enhanced the README documentation to reflect these new features and their usage, ensuring users can easily configure memory settings.

* refactor: improve memory store abstraction in voice assistant

- Introduced a new MemoryStore base class to enhance the memory management structure.
- Updated MainControlExtension to utilize MemoryStore instead of directly using PowerMemSdkMemoryStore.
- Renamed parameters in memory search methods for consistency and clarity, changing 'category_query' to 'query'.
- Revised README documentation to reflect changes in memory retrieval parameters and class structure.

* feat: update personalized greeting template in voice assistant

- Enhanced the PERSONALIZED_GREETING_TEMPLATE to include a condition for responding in the most commonly used language of the user's location/region if such information is available in the memory summary.
- This change aims to improve user engagement by making greetings more contextually relevant and personalized.

* feat: refine personalized greeting template in voice assistant

- Updated the PERSONALIZED_GREETING_TEMPLATE to emphasize the assistant's friendly and helpful nature, enhancing user interaction.
- This change aims to create a more engaging and contextually relevant greeting experience for users.

* fix: correct typographical errors in prompt templates for voice assistant

- Updated the CONTEXT_MESSAGE_WITH_MEMORY_TEMPLATE and PERSONALIZED_GREETING_TEMPLATE to fix typographical inconsistencies, ensuring proper use of apostrophes.
- This change enhances the clarity and professionalism of the prompt strings used in the extension.

* feat: add user memory management to voice assistant

- Introduced `enable_user_memory` configuration option to control user-specific memory functionality.
- Updated `MainControlExtension` to initialize the appropriate memory store based on the new configuration.
- Added `PowerMemSdkUserMemoryStore` class for handling user-specific memory operations.
- Revised documentation to include the new configuration parameter and its implications for memory management.

* refactor: update type hinting for conversation parameter in memory store classes

- Changed type hint for the `conversation` parameter from `List[dict]` to `list[dict]` in the `add` method across all memory store classes for consistency with Python's type hinting conventions.
- This change improves code clarity and aligns with modern Python practices.

* docs: enhance README documentation for voice assistant with PowerMem

- Updated README files to include links to the PowerMem repository, improving accessibility to relevant resources.
- Clarified the description of PowerMem's memory management capabilities in the context of the voice assistant, enhancing user understanding of its features.

* feat: add user memory configuration options to manifest

- Introduced new configuration options in the manifest.json for user memory management, including `enable_user_memory`, `memory_save_interval_turns`, and `memory_idle_timeout_seconds`.
- These additions enhance the flexibility and control over user-specific memory functionalities in the voice assistant.

* feat: implement user profile retrieval in memory store

- Added a new abstract method `get_user_profile` to the MemoryStore class for fetching user profiles.
- Implemented the `get_user_profile` method in the PowerMemSdkMemoryStore class to retrieve and format user profile information.
- Updated the MainControlExtension to utilize the new method for accessing user memory summaries, enhancing the assistant's ability to provide personalized interactions.

* fix: correct user memory initialization logic in MainControlExtension

- Updated the condition for initializing the memory store in MainControlExtension to ensure that the PowerMemSdkMemoryStore is used when user memory is disabled, aligning with the intended configuration behavior.
- This change improves the accuracy of memory management based on user settings.

* chore: remove .gitignore file for voice assistant with PowerMem

- Deleted the .gitignore file from the voice assistant example directory, which previously excluded the 'tenapp/data/' directory. This change simplifies the project structure by removing unnecessary files.

* fix: update language configuration in voice assistant property file

- Modified the language parameter in the property.json file to specify only "en-US", removing "zh-CN". This change streamlines language support for the voice assistant, ensuring it operates with a single language setting.

* fix: update dependencies and configuration in voice assistant files

- Downgraded versions and updated hashes for several dependencies in manifest-lock.json to ensure compatibility.
- Changed the STT addon from "azure_asr_python" to "deepgram_asr_python" and updated its parameters in property.json for improved speech recognition.
- Switched the TTS addon from "azure_tts_python" to "elevenlabs_tts2_python" and modified its configuration for better performance and output quality.

* fix: enhance memory store initialization error handling in MainControlExtension

- Added error handling during the initialization of the PowerMem memory store to log failures and allow the extension to continue without memory functionality if initialization fails.
- Improved logging to provide detailed information about the initialization process and any exceptions encountered, enhancing debugging capabilities.

* fix: update logging in PowerMemSdkUserMemoryStore for clarity

- Changed log messages in the PowerMemSdkUserMemoryStore class to reflect the correct class name, enhancing clarity in log outputs.
- This update ensures that logs accurately represent the source of memory operations, aiding in debugging and monitoring.

* fix: improve logging and error handling in user profile retrieval methods

- Enhanced logging in the `get_user_profile` and `get_user_profile_by_query` methods of both `PowerMemSdkMemoryStore` and `PowerMemSdkUserMemoryStore` classes to provide detailed information about the retrieval process and results.
- Added error handling to log exceptions and tracebacks, improving the robustness of user profile retrieval and aiding in debugging efforts.

* fix: simplify user profile retrieval in MainControlExtension

- Removed the query parameter from the `get_user_profile` method call in MainControlExtension, allowing for a more straightforward retrieval of user memory summaries.
- This change enhances the clarity and efficiency of the memory retrieval process.

* fix: refine user profile retrieval parameters in MainControlExtension

- Updated the `get_user_profile` method call in MainControlExtension to include user_id and agent_id parameters, enhancing the specificity of user memory retrieval.
- This change improves the accuracy of the memory summary fetched for the user, ensuring personalized interactions.

* fix: remove agent_id parameter from user profile retrieval in PowerMemSdkUserMemoryStore

- Eliminated the agent_id parameter from the profile retrieval call in the PowerMemSdkUserMemoryStore class, streamlining the user profile fetching process.
- This change enhances the clarity and efficiency of user memory interactions, aligning with recent updates to improve memory retrieval methods.

* fix: update configuration settings in voice assistant README

- Added timezone configuration for better localization support.
- Renamed database configuration parameters for clarity and consistency.
- Introduced OceanBase and PostgreSQL configuration settings for enhanced database integration.
- Added optional graph store configuration to support knowledge graph functionality.
- Improved overall documentation for better user guidance on setup and configuration.

* fix: enable user memory functionality in voice assistant configuration

- Updated configuration files to enable user memory and memorization features for the voice assistant.
- Adjusted property.json and config.py files to reflect the changes, enhancing user profile management capabilities.
- Improved README documentation to clarify the new settings and their implications for user interactions.

* fix: update Dockerfile and manifest versions for voice assistant

- Updated the Dockerfile to use the latest builder image version 0.7.12.
- Changed the version specification for the ten_ai_base dependency in manifest.json files from "=0.7.22" to "0.7" to allow for more flexible version compatibility.

* feat: add Docker support for voice assistant with PowerMem

- Introduced a Dockerfile for building the voice assistant application.
- Added a docker-compose.yml file to manage services, including the voice assistant and SeekDB database.
- Created a .gitignore file to exclude data directory from version control.

* feat: update configuration for voice assistant with PowerMem

- Revised README files to enhance clarity on required and optional environment variables for the voice assistant.
- Changed the default database provider from SQLite to OceanBase in configuration templates.
- Removed unnecessary comments and default settings to streamline the configuration process.

* feat: enhance README for voice assistant with Docker instructions

- Added a new section in the README for Docker setup, including commands to start the container and access the application.
- Improved clarity on the quick start process for users deploying the voice assistant.

* docs: expand README for voice assistant with detailed setup instructions

- Enhanced the README with a comprehensive table of contents for easier navigation.
- Added detailed sections on prerequisites, configuration, and verification steps for both Docker and local development setups.
- Improved clarity on required environment variables and provided examples for configuration.
- Included troubleshooting tips and common issues to assist users in resolving potential setup problems.

* feat: add Chinese documentation for voice assistant with PowerMem

- Created a new README_CN.md file to provide comprehensive setup instructions and information in Chinese.
- Updated the main README.md to include a link to the Chinese documentation for better accessibility for Chinese-speaking users.

* refactor: remove Chinese README for voice assistant with PowerMem

- Deleted the README_CN.md file as it is no longer needed.
- Updated the main README.md to remove the link to the Chinese documentation, streamlining the content for English-speaking users.

* docs: update README for voice assistant with PowerMem

- Modified the "Quick Start" section header in the README to improve clarity by replacing the hyphen with three dashes.
- Ensured consistency in formatting for better readability.

* docs: add configuration guide for voice assistant with PowerMem

- Introduced a new CONFIGURATION.md file detailing environment variables and configuration options for the voice assistant.
- Updated README.md to include references to the new configuration guide, enhancing user support for setup and alternative providers.

* chore: update configuration and documentation for voice assistant with PowerMem

- Enhanced .gitignore to include .env file and data directory for better environment management.
- Updated docker-compose.yml to reference the local .env file for environment variables.
- Revised README.md to clarify steps for configuring environment variables and starting services, including new instructions for SeekDB container setup.
- Adjusted Taskfile.yml to use the local .env file for dotenv configuration.

* feat: update Dockerfile and requirements for voice assistant with PowerMem

- Added a step in the Dockerfile to copy and install Python dependencies from requirements.txt.
- Updated requirements.txt to specify a minimum version for the powermem package, ensuring compatibility.

* docs: remove outdated environment configuration from README for voice assistant with PowerMem

- Eliminated the detailed environment configuration section from README.md to streamline documentation.
- Focused on providing a concise overview of configuration options, enhancing clarity for users.

* docs: simplify service startup instructions in README for voice assistant with PowerMem

- Reorganized the steps for starting services with Docker Compose to improve clarity.
- Renumbered the steps to create a more logical flow, enhancing user experience.

* feat: add .env.example file and update README for voice assistant with PowerMem

- Introduced a new .env.example file containing environment variable configurations for the voice assistant.
- Updated README.md to reflect changes in environment setup instructions, including a more streamlined process for copying the .env.example file.
- Renumbered steps in the README to improve logical flow and clarity for users.

* docs: update README to remove unnecessary environment variable URLs for voice assistant with PowerMem

- Removed LLM_BASE_URL and EMBEDDING_BASE_URL from the README.md as they are no longer needed.
- Streamlined the documentation to focus on essential configuration options, enhancing clarity for users.

* docs: update .env.example and README for voice assistant with PowerMem

- Added new options for LLM_PROVIDER and EMBEDDING_PROVIDER in both .env.example and README.md to include ollama, vllm, anthropic, deepseek, huggingface, and lmstudio.
- Ensured consistency between the environment variable documentation and the actual configuration options available.

* feat: enhance PowerMem integration in voice assistant example

- Updated .env.example to use localhost for OCEANBASE_HOST.
- Added PowerMem configuration options in property.json for vector store, LLM, and embedder.
- Modified MainControlConfig to include powermem_config.
- Updated memory store initialization to utilize PowerMem configuration.
- Improved logging for configuration loading in extension.py.

* chore: remove obsolete files and update Dockerfile for voice assistant with PowerMem

- Deleted .gitignore and docker-compose.yml as they are no longer needed.
- Updated Dockerfile to copy .env.example to .env and added seekdb as a build stage.
- Modified Taskfile.docker.yml to include a new task for running the seekdb server in production.
- Adjusted Taskfile.yml to reference the correct .env path.

* refactor: remove .env.example and streamline README for voice assistant with PowerMem

- Deleted obsolete .env.example file to simplify configuration.
- Updated README.md to consolidate PowerMem configuration details and quick start instructions.
- Enhanced clarity on environment variable setup and added new configuration options for memory management in the extension.
- Improved documentation on memory functionality and user greeting generation.

* docs: update README for voice assistant with PowerMem to include seekdb server setup

- Added instructions to start the seekdb server as the first step in the quick start guide.
- Renumbered subsequent steps for clarity and consistency.
- Enhanced overall documentation for better user guidance.

* docs: update README for voice assistant with PowerMem to include ROOT_PASSWORD environment variable

- Added ROOT_PASSWORD environment variable to the seekdb server setup instructions for clarity.
- Ensured consistency with previous updates to the README for better user guidance.

* docs: refine README for voice assistant with PowerMem to enhance clarity on environment variables

- Removed comments from the .env section for a cleaner presentation.
- Ensured consistency in formatting for LLM and embedding provider sections.
- Improved overall readability of the configuration instructions.

* docs: remove CONFIGURATION.md for voice assistant with PowerMem

- Deleted the outdated configuration guide to streamline documentation.
- Consolidated environment variable and configuration details into the README for improved clarity and accessibility.

* docs: Update database name in README for PowerMem

---------

Co-authored-by: Ethan Zhang <[email protected]>

* fix: calculate start time for ASR transcript

* fix: version update for sarvam asr

* feat: support tman install --locked

* feat: add more test cases

* fix: refine code

* fix: revert changes

* feat: support tracing in tman

* feat: add more tracing point in tman

* chore: add more debug info

* chore: add more debug info

* fix: refine codes

* fix: move stack size setting from runtime to build time

---------

Co-authored-by: Hu Yueh-Wei <[email protected]>

* feat: support opentelemetry

* fix: refine code

* feat: add metrics for lifecycle timecost

* fix: refine code

* fix: add metric for cmd processing time cost

* fix: refine code

* fix: refine code

* fix: refine code

* feat: add telemetry deployment configs

* fix: refine code

* fix: refine code

* fix: refine codes

* fix: refine codes

* fix: refine codes

* fix: merge host and port to endpoint in prometheus config

* fix: refine code

* fix: refine code

---------

Co-authored-by: Hu Yueh-Wei <[email protected]>

…sage formatting (#1902)

- Replaced the deprecated `from_utf8_lossy` function with `String::from_utf8_lossy` for improved clarity and consistency in the `PtyMessage` display implementation.
- This change enhances code readability and aligns with current Rust best practices.