Agent skill
fix-imports
Repair broken imports automatically across the project
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/data/fix-imports
SKILL.md
Fix Broken Imports
I'll systematically fix import statements broken by file moves or renames, with full continuity across sessions.
Arguments: $ARGUMENTS - specific paths or import patterns to fix
Token Optimization Strategy
Target: 70% reduction (2,500-3,500 → 750-1,050 tokens)
Core Principles:
- Grep-First Detection - Search for import errors before reading files
- Git-Diff Scoping - Only analyze recently changed files
- Template-Based Patterns - Use cached import resolution strategies
- Early Exit - Stop if no broken imports detected
- Bash-Based Updates - Use sed/awk for simple path corrections
Optimization Patterns:
Phase 1: Smart Detection (200-300 tokens)
# 1. Early exit check - Use language server diagnostics if available
mcp__ide__getDiagnostics # Get import errors from IDE
# 2. Git diff scoping - Only check changed files
git diff --name-only HEAD~5 # Recent changes only
# 3. Grep for import statements with errors
Grep "^import.*from ['\"].*['\"]" --glob "*.{js,ts,jsx,tsx}"
Grep "^from .* import" --glob "*.py"
Phase 2: Cached Resolution (100-200 tokens)
- Check
.claude/cache/fix-imports/patterns.jsonfor:- Common path transformations (e.g.,
components/→@/components/) - Recent file moves and renames
- Framework-specific import patterns
- Common path transformations (e.g.,
- Reuse resolution strategies from previous sessions
- Skip re-detection of project structure
Phase 3: Targeted Fixes (300-400 tokens)
# Use Bash for simple path corrections
sed -i 's|from "../components/|from "@/components/|g' file.ts
# Use Edit only for complex updates
Edit file.ts "old_import" "new_import"
Phase 4: Build Verification (150-250 tokens)
# Quick syntax check instead of full build
npx tsc --noEmit --skipLibCheck # TypeScript
eslint --fix file.ts # JavaScript
python -m py_compile file.py # Python
Token Budget Breakdown:
- Detection: 200-300 tokens (Grep + git diff)
- Resolution: 100-200 tokens (cache lookup)
- Fixes: 300-400 tokens (Bash + Edit)
- Verification: 150-250 tokens (build tools)
- Total: 750-1,150 tokens (70% reduction from 2,500-3,500)
Caching Strategy:
- Import patterns cache: Framework-specific import conventions
- File location cache: Moved/renamed file mappings
- Resolution strategy cache: Successful fix patterns
- Cache invalidation: On project structure changes
Progressive Disclosure:
- Quick scan → Exit if no issues (50 tokens)
- Targeted analysis → Focus on error locations (200 tokens)
- Incremental fixes → One import at a time (300 tokens)
- Full verification → Only if needed (200 tokens)
Session Resume Optimization:
- Load
fix-imports/state.json→ Skip completed fixes - Resume from last unresolved import
- Reuse cached resolution patterns
- Savings: 70% on resumed sessions
Optimization Status: ✅ Fully Optimized (Phase 2 Batch 3D-F, 2026-01-26)
- Achieves 70% token reduction target
- Maintains accuracy and safety
- Preserves full session continuity
Caching Behavior:
- Session location:
fix-imports/(plan.md, state.json) - Cache location:
.claude/cache/fix-imports/ - Caches: Import patterns, file locations, resolution strategies
- Cache validity: Until session completed
- Shared with:
/scaffold,/refactorskills
Session Intelligence
I'll maintain import fixing progress:
Session Files (in current project directory):
fix-imports/plan.md- All broken imports and fixesfix-imports/state.json- Resolution progress
IMPORTANT: Session files are stored in a fix-imports folder in your current project root
Auto-Detection:
- If session exists: Resume from last import
- If no session: Scan for broken imports
- Commands:
resume,status,new
Phase 1: Import Analysis
MANDATORY FIRST STEPS:
- Check if
fix-importsdirectory exists in current working directory - If directory exists, check for session files:
- Look for
fix-imports/state.json - Look for
fix-imports/plan.md - If found, resume from existing session
- Look for
- If no directory or session exists:
- Scan for all broken imports
- Create fix plan
- Initialize progress tracking
- Show import issues summary
Note: Always look for session files in the current project's fix-imports/ folder, not ../../../fix-imports/
I'll detect broken imports:
Import Patterns:
- File not found errors
- Module resolution failures
- Moved or renamed files
- Deleted dependencies
- Circular references
Smart Detection:
- Language-agnostic scanning
- Path alias understanding
- Barrel export recognition
- External vs internal imports
Phase 2: Resolution Planning
Based on analysis, I'll create resolution plan:
Resolution Strategy:
- Exact filename matches
- Similar name suggestions
- Export symbol search
- Path recalculation
- Import removal if needed
I'll write this plan to fix-imports/plan.md with:
- Each broken import location
- Possible resolutions
- Confidence level
- Fix approach
Phase 3: Intelligent Fixing
I'll fix imports matching your patterns:
Resolution Patterns:
- Update relative paths correctly
- Maintain path alias usage
- Preserve import grouping
- Follow sorting conventions
Ambiguity Handling:
- Show multiple matches
- Provide context for choice
- Never guess when uncertain
- Track decisions for consistency
Phase 4: Incremental Fixing
I'll fix imports systematically:
Execution Process:
- Create git checkpoint
- Fix import with verification
- Check for new breaks
- Update plan progress
- Move to next import
Progress Tracking:
- Mark each fix in plan
- Record resolution choices
- Create meaningful commits
Phase 5: Verification
After fixing imports:
- Syntax validation
- No new broken imports
- Circular dependency check
- Build verification if possible
Context Continuity
Session Resume:
When you return and run /fix-imports or /fix-imports resume:
- Load broken imports list
- Show fixing statistics
- Continue from last import
- Apply same resolution patterns
Progress Example:
RESUMING IMPORT FIXES
├── Total Broken: 34
├── Fixed: 21 (62%)
├── Current: src/utils/helpers.js
└── Next: src/components/Header.tsx
Continuing fixes...
Practical Examples
Start Fixing:
/fix-imports # Fix all broken imports
/fix-imports src/ # Focus on directory
/fix-imports "components" # Fix component imports
Session Control:
/fix-imports resume # Continue fixing
/fix-imports status # Check progress
/fix-imports new # Fresh scan
Safety Guarantees
Protection Measures:
- Git checkpoint before fixes
- Incremental changes
- Verification after each fix
- Clear decision audit
Important: I will NEVER:
- Guess ambiguous imports
- Break working imports
- Add AI attribution
- Create circular dependencies
What I'll Actually Do
- Scan completely - Find all broken imports
- Analyze smartly - Understand move patterns
- Fix accurately - Correct paths precisely
- Track thoroughly - Perfect continuity
- Verify always - Ensure imports work
I'll maintain complete continuity between sessions, always resuming exactly where we left off with consistent resolution patterns.
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
Didn't find tool you were looking for?