Lint Fix Loop

Iteratively fix linting errors until the codebase is clean.

Philosophy: Fix errors properly, don't suppress them. Understand why the rule exists.

The Job

1. •Run lint command to see all errors
2. •Fix ONE error at a time
3. •Run lint again to verify fix didn't break anything
4. •Repeat until zero errors

Iteration Flow

┌─────────────────────────────────────────────────┐
│ 1. Run lint command                             │
│ 2. If zero errors → COMPLETE                    │
│ 3. Pick ONE error to fix                        │
│ 4. Understand why the rule exists               │
│ 5. Fix the error properly                       │
│ 6. Run lint again to verify                     │
│ 7. Loop back to step 1                          │
└─────────────────────────────────────────────────┘

Fix Priority

Fix errors in this order:

1. •Type errors - These often cause other errors
2. •Import errors - Missing/wrong imports
3. •Unused variables - Dead code
4. •Style errors - Formatting, naming
5. •Complexity warnings - Refactoring needed

Fixing Guidelines

DO:

• •Understand why the lint rule exists
• •Fix the root cause
• •Check if the fix is consistent with codebase patterns
• •Run tests after fixing

DON'T:

• •Add // eslint-disable comments (except for rare valid cases)
• •Suppress errors without understanding them
• •Make changes that break functionality

Stop Condition

When lint command returns zero errors:

1. •Output: <promise>COMPLETE</promise>

If stuck on an error:

1. •Note why in progress.txt
2. •If it's a false positive, document and disable with comment
3. •If truly blocked, output: <promise>BLOCKED</promise>

Example Usage

User: "Fix all lint errors"

The skill will:

1. •Run lint command
2. •Fix errors one by one
3. •Report when all errors are fixed

Checklist Per Iteration

• • Ran lint command
• • Identified ONE error to fix
• • Understood why the rule exists
• • Fixed the error properly (not suppressed)
• • Verified fix with another lint run
• • Verified tests still pass