Configuration

CognitiveLint uses cosmiconfig for config discovery. It searches for (in order):

• cognitivelint key in package.json
• .cognitivelintrc
• .cognitivelintrc.json
• .cognitivelintrc.yaml
• .cognitivelintrc.yml
• .cognitivelintrc.js
• .cognitivelintrc.mjs
• cognitivelint.config.js
• cognitivelint.config.mjs
• cognitivelint.config.ts

Full example

cognitivelint.config.js

export default {
  // Files to analyze
  include: ['src/**/*.tsx', 'src/**/*.jsx'],

  // Files to skip
  exclude: [
    '**/node_modules/**',
    '**/*.test.*',
    '**/*.spec.*',
    '**/*.stories.*',
    '**/components/ui/**',   // skip generated component libraries
  ],

  rules: {
    // Turn a rule off entirely
    'error-prevention/modal-nesting': 'off',

    // Change severity
    'feedback/missing-loading-state': { severity: 'critical' },

    // Change severity + pass options
    'cognitive-load/excessive-primary-actions': {
      severity: 'high',
      options: { maxPrimaryActions: 3 },
    },
    'cognitive-load/long-forms': {
      options: { maxFields: 10 },
    },
    'cognitive-load/filter-overload': {
      options: { maxFilters: 7 },
    },
    'cognitive-load/dense-tables': {
      options: { maxColumns: 10 },
    },
  },

  // CI thresholds (equivalent to CLI flags)
  minScore: 70,
  maxFindings: 50,
};

Rule IDs

Rule IDs follow the pattern {category}/{rule-name}:

cognitive-load/excessive-primary-actions
cognitive-load/long-forms
cognitive-load/filter-overload
cognitive-load/dense-tables
trust/unexplained-disabled
trust/ownership-ambiguity
trust/missing-ownership
feedback/missing-loading-state
feedback/missing-empty-state
feedback/missing-success-feedback
feedback/no-progress-indicator
discoverability/missing-search
discoverability/hidden-primary-action
discoverability/empty-navigation
consistency/inconsistent-button-labels
error-prevention/destructive-no-confirm
error-prevention/no-undo
error-prevention/modal-nesting
error-prevention/confirmation-fatigue

Rule severity values

Value	Effect
'off'	Disable the rule entirely
'low'	Finding deducts −2 points
'medium'	Finding deducts −5 points
'high'	Finding deducts −10 points
'critical'	Finding deducts −15 points

Scoring

Deductions are multiplied by the rule's confidence score (0–100%). A finding at 60% confidence at HIGH severity costs 0.6 × 10 = 6 points instead of 10.

Final score is clamped to 0–100.

Grade	Score
A	90–100
B	80–89
C	70–79
D	60–69
F	0–59