.claude/skills/confidence-check/SKILL.md
Pre-implementation confidence assessment to prevent wrong-direction work. Evaluates readiness across 5 weighted criteria before proceeding with implementation.
npx skillsauth add anyproto/anytype-swift confidence-checkInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Prevent costly wrong-direction work by assessing implementation readiness before starting. Spend 100-200 tokens on assessment to save 5,000-50,000 tokens on misdirected implementation.
Evaluate each criterion and calculate a weighted confidence score:
| Check | Weight | What to Verify |
|-------|--------|----------------|
| No Duplicates | 25% | Search codebase for existing functionality |
| Pattern Compliance | 25% | Aligns with MVVM, Coordinator, @Injected DI |
| Documentation Verified | 20% | Checked middleware, protobuf, existing services |
| Design System Applied | 15% | Uses Loc.*, Color.*, Image(asset:) |
| Root Cause Clear | 15% | Understand actual problem vs symptoms |
Score >= 90% --> PROCEED with implementation
Score 70-89% --> PAUSE - present alternatives, ask clarifying questions
Score < 70% --> STOP - request more context before proceeding
Before implementing, search for existing solutions:
# Search for similar functionality
rg "similar_function_name" --type swift
rg "RelatedClass" --type swift
Pass: No existing implementation found, or existing code needs extension Fail: Duplicate functionality exists that should be reused
Verify alignment with project architecture:
@Injected for dependency injection@MainActor on ViewModelsPass: Planned approach follows all relevant patterns Fail: Proposed solution violates established patterns
Confirm understanding of existing systems:
Pass: Have concrete evidence from code exploration Fail: Assumptions without code verification
Verify UI implementation uses project standards:
Loc.keyName (not hardcoded strings)Color.Text.primary, Color.Shape.* (not custom colors)Image(asset: .X24.iconName) (not SF Symbols)Pass: All UI elements use design system Fail: Custom colors, hardcoded strings, or non-standard icons
Ensure you understand the actual problem:
Pass: Can articulate problem and solution clearly Fail: Only treating symptoms or unclear on requirements
Task: Add a new settings toggle for notification preferences
CHECK 1 - No Duplicates: [PASS - 25%]
- Searched: rg "notification.*toggle" --type swift
- Found: No existing toggle, only NotificationService
CHECK 2 - Pattern Compliance: [PASS - 25%]
- Will use SettingsViewModel (existing)
- Will extend SettingsCoordinator for new screen
- Will inject NotificationService via @Injected
CHECK 3 - Documentation Verified: [PASS - 20%]
- Reviewed SettingsViewModel.swift
- Checked NotificationService protocol
- Confirmed UserDefaults storage pattern
CHECK 4 - Design System Applied: [PASS - 15%]
- Will use Loc.Settings.notifications
- Will use Toggle from DesignSystem
- Will use Color.Text.primary
CHECK 5 - Root Cause Clear: [PASS - 15%]
- Users need granular notification control
- Currently only global on/off exists
- Need per-category toggles
TOTAL SCORE: 100% --> PROCEED
For faster assessment, use this condensed format:
CONFIDENCE CHECK:
[ ] No duplicates found (25%)
[ ] Follows MVVM/Coordinator/DI patterns (25%)
[ ] Verified existing code/docs (20%)
[ ] Uses Loc/Color/Image design system (15%)
[ ] Root cause understood (15%)
Score: ___% --> [PROCEED/PAUSE/STOP]
Skip confidence check for:
| Scenario | Without Check | With Check | |----------|--------------|------------| | Duplicate implementation | 5,000+ tokens wasted | 100 tokens saved it | | Wrong architecture | 10,000+ tokens rework | 150 tokens caught early | | Missing context | 8,000+ tokens redoing | 200 tokens asked first |
Rule of thumb: If implementation will take >500 tokens, spend 100-200 on confidence check first.
Navigation: This skill integrates with ios-dev-guidelines and design-system-developer for pattern verification.
development
Smart router to testing patterns and practices. Use when writing unit tests, creating mocks, testing edge cases, or working with Swift Testing and XCTest frameworks.
development
Audit and improve SwiftUI runtime performance through code review and Instruments guidance. Use for diagnosing slow rendering, janky scrolling, excessive view updates, or layout thrash in SwiftUI apps.
development
SwiftUI view structure, composition, and best practices. Use when refactoring SwiftUI views, organizing view files, or extracting subviews.
development
Write, review, or improve SwiftUI code following best practices for state management, view composition, performance, macOS-specific APIs, and iOS 26+ Liquid Glass adoption. Use when building new SwiftUI features, refactoring existing views, reviewing code quality, or adopting modern SwiftUI patterns.