Server lifecycle:
- Introduce service.Service to manage HTTP + gRPC startup/shutdown
- Fix gRPC mutex deadlock in StartServer when port is occupied
- Add GracefulStop with 5s timeout before forced Stop in grpc.go
- Pre-check HTTP and gRPC port availability in cmd/start.go
- Print gRPC server address in startup access-points block
gRPC client refactor:
- Move token manager and client from tai/grpc/ to grpc/client/
- Add backward-compatible aliases in tai/yao.go and tai/token.go
- Update cmd/run.go to import grpc/client directly (no tai dependency)
Sandbox v2 docker migration:
- Delete sandbox/v2/docker/ (moved to tai repo)
- Update sandbox/docker/build.sh hint to point to tai repo
- Clean up .gitignore entries for removed docker directory
- Temporarily disable SandboxV2Test and BenchmarkSandboxV2 in CI
(docker images need rebuild after tai repo migration)
Tai integration:
- Add direct-mode registration API handlers in tai/api/
- Add heartbeat handler and token management wrappers
- Update tai/registry and tai/tunnel for latest protocol
- Replace yao-grpc references with tai call in docs
Made-with: Cursor
The bubbletea-based TUI for agent request visualization was not
practical and added significant binary size. This removes the TUI
entirely: the --tui flag, the agent context TUI model/messages, and
all SendTUI/GetTUIProgram call sites. Development-mode logging now
always prints ANSI-colored output to stdout.
Drops direct dependencies on charmbracelet/bubbletea and lipgloss.
Made-with: Cursor
- Introduce a new "tunnel" scheme for the Tai client, allowing connections through Yao's reverse proxy.
- Enhance the Tai registry to manage tunnel-connected nodes and their ports.
- Add WebSocket and reverse proxy routes for tunnel connections in the OpenAPI server.
- Implement tunnel-specific proxy and VNC handling to facilitate communication with containerized environments.
- Update gRPC environment variable handling to support tunnel connections.
These changes improve the Tai service's flexibility and connectivity options, enabling better integration with remote and containerized environments.
- Add TAI_TEST_HOST_IP environment variable to CI workflows for unit tests, allowing better connectivity to the gRPC server from Docker containers.
- Update the `run.go` file to parse command-line arguments correctly.
- Modify the test utility to return the gRPC address reachable from Docker, improving integration test reliability.
- Refactor integration tests to utilize the new relay address function, ensuring proper communication with the Yao gRPC server.
These changes improve the testing framework and enhance the configuration for better service interaction during CI runs.
- Improve health check logic in CI workflows for both HTTP and gRPC readiness of the Tai service, ensuring clearer error reporting if the service fails to start.
- Update the OAuth Device Flow implementation to support additional claims during device authorization, enhancing the flexibility of the authorization process.
- Refactor the `AuthorizeDevice` method to accept extra claims, allowing for more detailed user context during authorization.
- Introduce a new utility function to extract bearer tokens from requests, streamlining token handling across the OpenAPI service.
These changes enhance the robustness of service readiness checks and improve the OAuth device authorization flow, contributing to a more reliable and flexible authentication mechanism.
- Add gRPC server configuration to the application, allowing for gRPC communication.
- Introduce new Makefile targets for gRPC unit testing and proto code generation.
- Update CI workflows to include gRPC tests with SQLite as the transport layer.
- Refactor the sandbox design to support multi-node capabilities and improve isolation.
- Enhance the service layer to facilitate internal request forwarding for gRPC APIs.
This commit lays the groundwork for integrating gRPC into the Yao SDK, improving performance and scalability.
- Add a `--force` flag to the `push` command in both agent and MCP modules, allowing users to overwrite existing versions by deleting them before pushing.
- Update the `PushOptions` struct to include the `Force` field, ensuring the functionality is integrated into the push logic.
- Enhance E2E tests to validate the behavior of the force push feature, ensuring that existing versions are deleted as expected when the flag is used.
- Introduce tests for packing directories to respect `.yaoignore` rules, improving file exclusion handling during the packaging process.
- Introduce new MCP and Robot command groups in the CLI, allowing for better organization of package management commands.
- Add corresponding command descriptions for MCP and Robot functionalities to improve user guidance.
- Update the agent command group to include additional commands for enhanced agent management.
- Modify .gitignore to exclude specific design markdown files from version control.
- Add environment variable for test application path in GitHub workflows to streamline testing setup.
- Add new indirect dependencies including various Charmbracelet packages for improved UI handling.
- Enhance logging in the Assistant module by adding tool completion and start logging for better traceability of tool calls.
- Modify context handling in the RequestLogger to support a stack-based assistant ID management, improving the logging structure for agent requests.
- Implement event service integration for better trace management and debugging capabilities.
- Update the output for API and OpenAPI endpoints in `cmd/start.go` to conditionally display the correct URL based on the OpenAPI server status.
- Remove redundant OpenAPI mode information display when the OpenAPI server is enabled, streamlining the API list output.
- Remove unnecessary timing and progress callback logic from the start command in `cmd/start.go`.
- Update the output message for the admin URL to reflect a change to the dashboard URL.
- Introduce a new tools configuration in `widgets/app/app.go` to report the availability of external tools like FFmpeg, Docker, and others, enhancing the application's tool inspection capabilities.
- Add external tool inspection functionality in the engine, allowing detection of tools like FFmpeg, PDF converters, and Docker.
- Update the `inspect` command to include detected tools in the output.
- Remove the studio configuration from the application, simplifying the configuration structure and related code.
- Clean up unused studio-related code in various files, including login processing and configuration management.
- Update start command to check if the current directory is a Yao app root or a subdirectory, providing clearer error messages for users.
- Modify installation logic to handle empty directories more effectively, ensuring the init app is installed only when appropriate.
- Improve welcome message formatting and update links for documentation and community resources.
Co-authored-by: Cursor <cursoragent@cursor.com>
- Update Makefile to include new production and UPX build targets for optimized binaries.
- Add BUILDOPTIONS constant to share/const.go for tracking build configurations.
- Modify version command to display build options.
- Update Dockerfile to install UPX for compression during builds.
- Add release and prod targets to streamline artifact creation.
Co-authored-by: Cursor <cursoragent@cursor.com>
- Updated README files to include new commands for running tests and extracting results for review.
- Introduced the `yao agent extract` command to facilitate extraction of test results from JSONL files into Markdown or JSON formats.
- Enhanced the `FormatAvailableResources` function to support localization and detailed information for agents and MCP tools.
- Improved output formatting for better readability and usability in test result documentation.
- Introduced support for OpenAPI mode, allowing dynamic routing and OAuth guards for API endpoints.
- Updated API root path handling to accommodate OpenAPI configurations, ensuring consistent URL structures.
- Added well-known routes for Yao metadata and OAuth discovery, improving API discoverability.
- Refactored middleware and guards to streamline OpenAPI integration, enhancing overall service functionality.
- Removed deprecated agent TypeScript file, simplifying the codebase and improving maintainability.
- Introduced the IWatchDirs interface to support watching multiple directories for changes, allowing templates to specify directories dynamically.
- Updated the watch command to handle multiple directories and display watched directories in the output.
- Added functionality to write locale files from a page's __locales directory to the public directory, improving localization support.
- Enhanced error handling during directory watching and locale file processing for better reliability and logging.
- Updated the `build` and `watch` commands to accept a single argument for `<sui>` and made the `<template>` argument optional, improving usability.
- Introduced a default template assignment for the `agent` SUI, ensuring a more intuitive setup for users.
- Enhanced error messages for command usage to provide clearer guidance on expected input format.
- Added new OpenAPI file to the bindata, improving the framework's capabilities for API integration.
- Introduced dynamic testing capabilities, allowing for multi-turn conversations with checkpoints through the new `DynamicRunner`.
- Enhanced the `runSingleTest` method to support dynamic mode, including the execution of before/after scripts and detailed output for dynamic test results.
- Added new output methods in `output.go` for dynamic test start, turns, checkpoints, and results, improving console feedback during testing.
- Updated `DESIGN_V2.md` and `TODO_V2.md` to reflect the new dynamic mode features, including simulator configurations and checkpoint definitions.
- Revised the `Case` struct in `types.go` to include fields for dynamic testing, such as `Simulator`, `Checkpoints`, and `MaxTurns`.
- Added support for generating test cases using agents and scripts through the new `LoadFromAgent` and `LoadFromScript` methods in the loader.
- Enhanced the `RunTests` method to handle different input sources, including agent-driven and script-based test case generation.
- Introduced a `--dry-run` flag to allow users to preview generated test cases without executing them.
- Updated relevant documentation in DESIGN_V2.md and TODO_V2.md to reflect the new features and implementation status.
- Updated DESIGN_V2.md to reflect changes in the handling of before/after scripts, removing the `scripts:` prefix and clarifying their usage in JSONL test cases.
- Enhanced runner.go to integrate global before/after hooks, ensuring they execute correctly before and after test cases.
- Revised types.go to include new fields for before/after scripts in test case and options structures.
- Improved TODO_V2.md to track the implementation progress of before/after script functionality and related tasks.
- Added utility function LoadAgentTestScripts to facilitate loading of test scripts from the agent's src directory.
- Introduced support for file attachments in test inputs using the `file://` protocol, allowing images, audio, and documents to be loaded and converted to appropriate formats.
- Updated `ParseInput` and related functions to handle file references, ensuring seamless integration of file content into messages.
- Enhanced error handling and path resolution for file loading, considering both relative paths and the `YAO_ROOT` environment variable.
- Expanded documentation to include examples of file attachments and their usage in test cases, improving clarity for users.
- Introduced a new script testing mode to allow testing of agent handler scripts (hooks, tools, etc.) using a Go-like interface, enabling better unit testing of TypeScript/JavaScript code.
- Enhanced the `LoadScripts` function to skip test files during script loading, ensuring only relevant scripts are processed.
- Refactored the test context creation to support custom context configurations via a JSON file, allowing for flexible authorization and metadata management during tests.
- Updated the test runner to handle script tests, including the ability to filter tests using regex patterns and manage custom context data.
- Improved documentation to include details on script testing usage, input formats, and available assertions, enhancing developer experience and clarity.
- Added a `noop` check in multiple logging methods (`LLMComplete`, `ToolStart`, `ToolComplete`, `HookStart`, `HookComplete`, and `HistoryLoad`) to prevent logging when the logger is in no-operation mode.
- Improved command handling in `root.go` by removing minimum argument requirements for commands and providing help output when no arguments are given.
- Introduced an `agent` command for better organization of agent-related functionalities in the CLI.
- Implemented automatic detection of the application root directory in `run.go` to streamline the application startup process.
- Cleaned up debug print statements in `config.go` to reduce clutter in the output.
- Introduced a `loadStep` function to wrap loading operations, providing timing and progress callback functionality.
- Updated the `Load` function to utilize `loadStep` for various components, improving visibility into loading durations.
- Enhanced the `start` command to display loading progress and duration in development mode, improving user experience during application startup.
- Refactored model loading functions to return models without migration, streamlining the loading process and improving performance.
- Updated the `printMCPs` function to enhance the separation of agent and standard MCPs by utilizing the Type field for better clarity.
- Improved error handling and client information retrieval, ensuring accurate display of transport types and labels for each MCP.
- Modified the MCP loading process to specifically load agent clients, streamlining the management of different client types.
- Enhanced output formatting to include tools count only for process transport, improving the readability of MCP information.
- Introduced a new `printMCPs` function to display loaded MCP clients, categorizing them into standard and agent clients for better clarity.
- Enhanced the `Load` function to include loading MCP clients from the assistants directory, improving the overall MCP management.
- Added tests for loading assistant MCPs, ensuring that the functionality works as expected and that agent clients are correctly identified and logged.
- Updated model loading to support agent models from assistants, ensuring proper integration and functionality across the system.
Remove deprecated studio package and refactor agent integration
- Deleted the studio package, which is no longer in use, to streamline the codebase.
- Updated references in the agent and chat modules to utilize the new agent package instead of the deprecated neo package.
- Ensured that all related middleware and routing functionalities are now aligned with the agent architecture, enhancing overall system coherence.
- Updated error messages to use fmt.Errorf with %s for better clarity and consistency.
- Enhanced logging statements to utilize formatted strings for improved readability.
- Refactored multiple Load functions across different modules to standardize error handling practices.
- Added CUI commit information to the Makefile for artifact generation.
- Enhanced version command output to include Yao and CUI commit details with improved formatting.
- Updated constant names in share/const.go for clarity on versioning.
- Fixed references to CUI commit in process and app files to align with new naming conventions.
- Added checks in various loading functions to ignore processing if the respective directories do not exist, improving error handling and preventing unnecessary operations.
- Updated the Load function signature to return warnings alongside errors, allowing for better feedback on loading issues across different components.
