serverpod_boost 0.1.2

AI acceleration for ServerPod development via MCP (Model Context Protocol). Provides AI assistants with deep context awareness of ServerPod projects, enabling high-quality code generation through sema [...]

Changelog

All notable changes to ServerPod Boost will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

0.1.2 - 2026-02-05

Fixed

• Install Command: Fixed skills not accessible when package is published to pub.dev
  • Moved skills from .ai/skills/ to lib/resources/skills/ for package accessibility
  • Updated _findBoostSkillsPath() to use package_config for path resolution
  • Skills are now properly bundled and accessible in published packages
• Dependencies: Added package_config: ^2.1.0 for package path resolution

0.1.1 - 2026-02-05

Fixed

• Install Command: Fixed skills not being copied to user projects
  • Corrected path calculation from .parent.parent.parent to .parent.parent
  • Added .ai/skills/ as package assets in pubspec.yaml
  • Skills are now properly copied during installation
• Install Command: Made interactive mode the default behavior
  • Added --non-interactive flag for silent installation
  • Removed --interactive flag (now default)
• Install Command: Added install executable alongside boost
  • Users can now run dart run serverpod_boost:install
  • Created bin/install.dart entry point
• Install Command: Fixed skill selection flow
  • Handles missing skills directory gracefully
  • Copies all built-in skills when selection list is empty
• Code Quality: Fixed all dart analyze warnings (41 warnings → 0)
  • Removed duplicate exports and unused imports
  • Fixed deprecated MCP imports
  • Replaced print() statements with stdout.writeln()
  • Fixed null safety issues and unnecessary casts
• Analysis Options: Removed undefined lint rule prefer_async_await

Changed

• Package URLs updated to use qazz92/serverpod_boost repository

0.1.0 - 2026-02-05

Added

• 20 MCP Tools for comprehensive ServerPod project analysis:
  • Tier 1 - Essential Tools: application_info, list_endpoints, endpoint_methods, list_models, list_skills, get_skill, model_inspector, config_reader, database_schema, migration_scanner
  • Tier 2 - Enhanced Tools: project_structure, find_files, read_file, search_code, call_endpoint, service_config, log_reader
  • Tier 3 - Database Tools: database_query
  • Tier 4 - CLI Tools: cli_commands
  • Tier 5 - Developer Tools: tinker
• MCP Server Implementation using package:mcp_server (v1.0.3) with JSON-RPC 2.0 over stdio
• Project Detection Capabilities for ServerPod v3 monorepos with automatic root, server, client, and Flutter app discovery
• YAML Parser for reading .spy.yaml protocol model definitions with full field extraction including types, persistence, and serialization settings
• Method Signature Parser for endpoint methods with return types, parameter names/types, and documentation extraction
• Service Locator (ServerPodLocator) for centralized project access and validation
• Database Query Tool for executing SQL queries against ServerPod databases
• CLI Commands Tool for executing ServerPod CLI commands (migrate, generate, etc.)
• Log Reader Tool for reading ServerPod server logs
• Tinker Tool for interactive code evaluation
• Comprehensive Test Coverage with integration tests

Implemented

• MCP protocol compliance with tools, resources, and prompts support
• Tool adapter layer for converting Boost tools to MCP server format
• Colored logging system with configurable verbosity
• Configuration management with local override support
• Error handling and validation throughout
• Migration framework from legacy MCP implementation to package:mcp_server

Tested

• Tool functionality across all 20 tools
• MCP server initialization and communication
• Project detection and validation
• YAML parsing for model definitions
• Method signature extraction
• Integration tests against real ServerPod projects