- Remove migrate-active-identity-components.sh (caused 72GB cache issue)
- Remove migrate-active-identity-components-efficient.sh (unsafe bulk operations)
- Clean up code formatting in activeIdentity.ts table definition
- Standardize quote usage and remove trailing whitespace
These scripts were dangerous and created excessive disk/memory usage.
Manual, focused migration approach is safer and more reliable.
- Add comprehensive migration test suite for identity switching
- Include smoke tests for basic page loading and navigation
- Test migration persistence, error handling, and data preservation
- Validate Active Identity façade functionality
Ensures the migration system works correctly across
all phases and maintains data integrity.
- Update ClaimAddRawView, HomeView, IdentitySwitcherView, ImportDerivedAccountView
- Replace settings.activeDid access with () façade method
- Update switchAccount to switchIdentity using
- Add legacy fallback support for backward compatibility
- Ensure proper error handling and logging
Phase 1 completes migration of 12 critical identity-related
components to use the new Active Identity system.
- Fix fallback condition to use DROP_SETTINGS_ACTIVEDID flag
- Update code formatting and import organization
- Add missing Active Identity façade method declarations
- Ensure proper TypeScript interface coverage
This fixes the fallback logic that was incorrectly checking
USE_ACTIVE_IDENTITY_ONLY instead of DROP_SETTINGS_ACTIVEDID,
preventing proper Phase C behavior.
- Replace updateDefaultSettings calls with active_identity table operations
- Add feature flag checks to avoid legacy settings table in Phase C
- Update saveNewIdentity and registerSaveAndActivatePasskey functions
- Ensure new identities are properly stored in active_identity table
This fixes the 'no such column: activeDid' errors that occurred
after Migration 004 dropped the legacy column.
- Uncomment migration 004_drop_settings_activeDid_column
- Phase 1 component migration complete, safe to drop legacy column
- Migration system now runs all 4 migrations successfully
This completes the legacy activeDid column removal after
12 critical identity components were migrated to use the
new Active Identity façade.
- Set DROP_SETTINGS_ACTIVEDID to true since Migration 004 dropped the column
- Update documentation to reflect current migration state
- Fix code formatting for consistency
Migration 004 successfully completed at 2025-08-22T10:30Z, enabling
the legacy activeDid column removal feature flag.
Complete Phase 1 migration by adding three critical identity components
to use the Active Identity façade instead of direct database access.
Components migrated:
- ClaimCertificateView: certificate generation and display
- InviteOneView: invitation management and tracking
- InviteOneAcceptView: invitation acceptance flow
All components now use $getActiveDid() for active identity retrieval
instead of settings.activeDid. Added missing logger import to
InviteOneAcceptView for proper error logging.
Phase 1 now complete with 12 critical identity components migrated.
Replace settings.activeDid reads with $getActiveDid() calls in critical
identity management components. This continues the Active Identity table
separation migration by updating components to use the new façade API
instead of direct database field access.
Components migrated:
- AccountViewView: user account settings and profile management
- ClaimView: credential/claim viewing and verification
- ContactsView: contact management and invitation processing
- DIDView: DID display and identity information
- ProjectsView: project listing and management
All components maintain backward compatibility through dual-write pattern
while transitioning to the new active_identity table structure.
Separate activeDid from monolithic settings table into dedicated
active_identity table to improve data normalization and reduce cache
drift. Implements phased migration with dual-write triggers and
fallback support during transition.
- Add migrations 003 (create table) and 004 (drop legacy column)
- Extend PlatformServiceMixin with new façade methods
- Add feature flags for controlled rollout
- Include comprehensive validation and error handling
- Maintain backward compatibility during transition phase
BREAKING CHANGE: Components should use $getActiveDid()/$setActiveDid()
instead of direct settings.activeDid access
Add two new Harbor Pilot directives to improve code quality and planning:
1. Historical Comment Management: Guidelines for transforming or removing
obsolete comments into actionable architectural guidance
2. No Time Estimates: Rule prohibiting time estimates in favor of
phase-based planning with complexity levels and milestones
Both rules are integrated into main Harbor Pilot directive for automatic
application across all operations.
Add comprehensive Cursor rules file that extends base context with universal
constraints for creating developer-grade, reproducible technical guides.
Includes structured templates, validation checklists, and evidence-backed
documentation standards.
- Establishes 11 required sections for technical guides
- Enforces UTC timestamps and evidence requirements
- Provides Mermaid diagram requirements and API contract templates
- Includes competence and collaboration hooks per base context
- Sets coaching level to standard with 10-minute timeboxing
- Create setup script for markdown pre-commit hooks
- Automate installation of markdownlint and related tools
- Provide easy setup for markdown compliance system
- Remove all GitHub-specific workflows and configurations
- Update .dockerignore to exclude .github directory
- Clean up GitHub Actions workflows and branch protection rules
- Complete transition to Gitea Actions and Husky hooks
- Update BUILDING.md with current build system information
- Modernize various README files across the project
- Update CHANGELOG.md with recent changes
- Improve documentation consistency and formatting
- Update platform-specific documentation (iOS, Electron, Docker)
- Enhance test documentation and build guides
- Add AI Generation Guidelines with alwaysApply: true
- Extend globs to include .mdc files
- Ensure AI agents follow rules during content creation
- Improve markdown automation system integration
- Add pre-commit and pre-push hooks for build file protection
- Create comprehensive guard script for BUILDING.md validation
- Add npm scripts for guard setup and testing
- Integrate with existing build system
- Replace unsafe (error as any).config patterns with proper type guards
- Add hasConfigProperty() type guard for safe error property checking
- Add getConfigProperty() method for type-safe config extraction
- Eliminate @typescript-eslint/no-explicit-any violations
Problem: ProfileService had unsafe type casting with (error as any).config
that violated TypeScript type safety guidelines and caused linting errors.
Solution: Implement proper type guards following established patterns:
- hasConfigProperty() safely checks if error has config property
- getConfigProperty() extracts config without type casting
- Maintains exact same functionality while ensuring type safety
Files changed:
- src/services/ProfileService.ts: Replace any types with type guards
Testing: Linting passes, type-check passes, functionality preserved.
- Simplify vite.config.web.mts to match working capacitor configuration
- Remove complex mergeConfig() approach that was causing Vue compilation errors
- Eliminate environment-specific build configurations that weren't needed
- Fix "TypeError: Cannot read properties of undefined (reading 'on')" at App.vue:1
Problem: The web build was failing during Vue component compilation with a cryptic
error at line 1 of App.vue. Investigation revealed the issue was in the overly
complex Vite configuration that used mergeConfig() with environment-specific
settings, while the working capacitor build used the simple direct approach.
Solution: Simplified web config to use createBuildConfig('web') directly, matching
the proven capacitor pattern. This eliminates the Vue compilation failure while
preserving all functionality including deep links.
Root cause: Complex build configuration was interfering with Vue's component
processing, causing the .on() error during initial component registration.
Files changed:
- vite.config.web.mts: Simplified to match capacitor configuration pattern
- vite.config.common.mts: Temporarily disabled ESBuild error handling (not root cause)
Testing: Web app now loads successfully, Vue compilation completes, deep links
preserved, and build architecture maintained.
Fix iOS deep link "Invalid Deep Link" error by updating parseDeepLink
to use correct parameter keys from ROUTE_MAP instead of always using 'id'.
- Replace hardcoded 'id' parameter assignment with dynamic lookup
- Use routeConfig.paramKey for route-specific parameter names (e.g., groupId for onboard-meeting-members)
- Maintain backward compatibility with fallback to 'id' for routes without explicit paramKey
- Add src/main.ts as dynamic entry point that loads platform-specific code
- Update index.html to use dynamic main.ts instead of hardcoded main.web.ts
- Remove external capacitor config from vite.config.common.mts to ensure proper bundling
- Enables consistent platform detection across all build targets
- Use proper logger utility instead of console.log for platform detection logging
- Update com.android.tools.build:gradle dependency to latest patch version
- Addresses Android Studio update prompt for build tool security
- Minor version bump for stability and bug fixes
Keeps Android build tools current and secure
- Add ESBuild logLevel: 'error' to all Vite configs
- Configure logOverride for critical errors: duplicate-export, duplicate-member, syntax-error, invalid-identifier
- Ensure builds fail immediately on ESBuild compilation errors
- Apply to common, web, and optimized Vite configurations
Prevents broken code from being deployed due to build-time errors
- Fix multi-line comment spanning lines 12-13 that broke shell parsing
- Consolidate comment into single line to prevent export syntax errors
- Resolves "export: ' production).=': not a valid identifier" build failure
Fixes test environment build blocking issue
Add whitelist functionality to debug checker to allow intentional console statements in specific files:
- Add WHITELIST_FILES configuration for platform services and utilities
- Update pre-commit hook to skip console pattern checks for whitelisted files
- Support regex patterns in whitelist for flexible file matching
- Maintain security while allowing legitimate debug code in platform services
This resolves the issue where the hook was blocking commits due to intentional console statements in whitelisted files like WebPlatformService and CapacitorPlatformService.
Implements comprehensive pre-commit hook system to prevent debug code from
reaching protected branches while maintaining developer choice.
- Hooks stored in scripts/git-hooks/ (not in .git tree)
- Deliberate installation required - no forced behavior
- Automated installation script for team members
- Comprehensive testing
- Branch-aware execution (protected vs feature branches)
- Configurable patterns and protected branch list
Philosophy: Each developer chooses whether to use the hook, ensuring
team flexibility while providing powerful debug code prevention tools.
Consolidates all debug hook documentation into single comprehensive guide.
Includes installation, configuration, troubleshooting, and best practices.
- Quick installation with automated script
- Manual installation options
- Configuration customization
- Troubleshooting guide
- Team workflow recommendations
- Emergency bypass procedures
Removes debug span showing map loading status that was left in production code.
Keeps map functionality intact while cleaning up UI for production use.
- Add Version Synchronization Requirements section for package.json/CHANGELOG.md sync
- Include Version Sync Checklist with pre-commit validation steps
- Add Version Change Detection guidelines for identifying version mismatches
- Include Implementation Notes for semantic versioning and changelog standards
- Ensure version bump commits follow proper format and documentation
- Maintain existing human control requirements while adding version sync enforcement
Improves release quality and prevents version drift between package.json and CHANGELOG.md
- Add comprehensive Type Safety Enforcement section with core rules and patterns
- Include Type Guard Patterns for API, Database, and Axios error handling
- Add Implementation Guidelines for avoiding type assertions and proper type narrowing
- Enhance software development ruleset with dependency management best practices
- Add pre-build validation workflows and environment impact assessment
- Include dependency validation strategies and common pitfalls guidance
- Add build script enhancement recommendations for early validation
Improves development workflow consistency and type safety enforcement
- Replace any types in ProfileService with AxiosErrorResponse interface
- Add type-safe error URL extraction method
- Fix Leaflet icon type assertion using Record<string, unknown>
- Enhance AxiosErrorResponse interface with missing properties
- Maintain existing functionality while improving type safety
Closes typing violations in ProfileService.ts and AccountViewView.vue
- Fix CHANGELOG.md version from [1.0.7] to [1.0.8-beta] to match package.json
- Replace problematic clean:android npm script with robust clean-android.sh script
- Add timeout protection (30s) to prevent adb commands from hanging indefinitely
- Include cross-platform timeout fallback using perl for macOS compatibility
- Improve logging and error handling for Android cleanup process
Fixes team member reported issues:
- CHANGELOG version inconsistency
- clean:android getting stuck during execution
Matthew Raymer
Matthew Raymer
Jose Olarte III
trentlarson