Agent Skills: Compliance Dashboard Skill

Real-time compliance tracking for kerala-rage kr-solidarity design system.

UncategorizedID: okgoogle13/careercopilot/compliance-dashboard

Install this agent skill to your local

pnpm dlx add-skill https://github.com/okgoogle13/careercopilot/tree/HEAD/.claude/skills/compliance-dashboard

Skill Files

Browse the full folder contents for compliance-dashboard.

Download Skill

Loading file tree…

.claude/skills/compliance-dashboard/SKILL.md

Skill Metadata

Name
compliance-dashboard
Description
Real-time compliance tracking for kerala-rage kr-solidarity design system.

Compliance Dashboard Skill

Purpose

Real-time compliance tracking CLI for kerala-rage kr-solidarity design system. Aggregates data from component specs, audit results, inventory analysis, and design token compliance to provide instant visibility into design system health.

Run the tool to get:

  • Overall compliance % (target: 80%)
  • Per-dimension metrics (typography, color, layout, [DEPRECATED_STYLE])
  • Component-level status (aligned vs needs refinement)
  • Prioritized work queues (which components to focus on next)

Quick Start

# Full report (human-readable)
python3 .claude/skills/compliance-dashboard/run.py

# JSON output (for CI/CD integration)
python3 .claude/skills/compliance-dashboard/run.py --json

# Single component analysis
python3 .claude/skills/compliance-dashboard/run.py --component LoginCard

# Component + JSON
python3 .claude/skills/compliance-dashboard/run.py --component LoginCard --json

Data Sources

The tool aggregates from multiple sources to compute component health:

  1. docs/design/component-batch-plan.yaml

    • Component metadata (name, mode, priority)
    • Target state (stage1, stage2, tests, tokens)
  2. docs/design/audits/<ComponentName>.json

    • Visual audit results (if exist)
    • Dimension scores (typography, color, layout, [DEPRECATED_STYLE])
    • Overall compliance score
  3. docs/design/generated/specs/

    • Component specifications (existence check)
  4. frontend/src/components/

    • Component code (existence check)
    • Located in components/<Name>/index.tsx or ui/<Name>.tsx
  5. frontend/src/components/tests/

    • Unit/integration tests (existence check)
  6. scripts/simple-component-inventory.sh --json

    • Hardcoded values detection (colors, spacing, shadows)
    • Duplicate component detection

Output Schema

JSON Mode (--json)

Complete structured output:

{
  "summary": {
    "overall_compliance": 0.78,
    "typography": 0.85,
    "color": 0.82,
    "layout": 0.74,
    "[DEPRECATED_STYLE]": 0.68,
    "migration": 0.92,
    "generated_at": "2026-02-10T02:07:37.749610+00:00"
  },
  "components": [
    {
      "name": "LoginCard",
      "mode": "new",
      "priority": "high",
      "scores": {
        "overall": null,
        "typography": null,
        "color": null,
        "layout": null,
        "[DEPRECATED_STYLE]": null
      },
      "status": "aligned|needs_refinement",
      "needs": [
        "needs_typography_fix",
        "needs_color_fix",
        "needs_layout_refine",
        "needs_botanical_integration",
        "needs_migration_cleanup",
        "needs_tests",
        "needs_stage2",
        "needs_spec"
      ],
      "has_spec": true,
      "has_mockup": false,
      "has_component_code": true,
      "has_tests": true,
      "hardcoded_values": false,
      "duplicate": false,
      "last_audit": "2026-02-10T00:00:00Z",
      "iterations": 2
    }
  ]
}

Human-Readable Output (default)

Kerala Rage Compliance Dashboard
================================
Generated: 2026-02-10T02:07:37

Overall Compliance: 78% ▲ (Target: 80%)
  Typography        85% ✓
  Color             82% ✓
  Layout            74% ▼
  [DEPRECATED_STYLE]         68% ▲
  Migration        100% ✓

Status: Approaching production-ready threshold

Top Components Needing Refinement:
  1. Lens (high priority)
     - needs_typography_fix
     - needs_tests
     - needs_migration_cleanup

  2. Jar (high priority)
     - needs_color_fix
     - needs_tests

  3. NewModule (high priority)
     - needs_stage2

Per-Component Logic

Status Derivation

A component is "aligned" when:

  • ✅ Spec exists (docs/design/generated/specs/<Name>.md)
  • ✅ Component code exists (frontend/src/components/<Name>/index.tsx)
  • ✅ Tests exist (if target.tests ≠ "none")
  • ✅ All dimension scores ≥ 0.8 (if audit exists)
  • ✅ No hardcoded values (if target.tokens = "strict")
  • ✅ Stage 2 complete (if target.stage2 = "complete")

Otherwise: "needs_refinement"

Needs Flags

Computed from gaps:

| Flag | Condition | |------|-----------| | needs_typography_fix | typography score < 0.8 | | needs_color_fix | color score < 0.8 | | needs_layout_refine | layout score < 0.8 | | needs_botanical_integration | [DEPRECATED_STYLE] score < 0.7 | | needs_migration_cleanup | mode = "migrate" AND hardcoded_values = true | | needs_tests | target.tests ≠ "none" AND no test file | | needs_stage2 | target.stage2 = "complete" AND (no mockup OR no code) | | needs_spec | spec file missing |

Dimension Scores

If audit JSON exists:

  • Score source: audit.dimensions.<dimension>.status
  • Mapping: pass=1.0, needs_refinement=0.7, fail=0.3
  • Alternative: audit.compliance_score / 100

If no audit JSON:

  • All scores: null
  • Iterations: 0
  • Last audit: null

Global Metrics

Overall Compliance

Percentage of components with status = "aligned"

overall_compliance = count(status = "aligned") / total_components

Per-Dimension Averages

Average score across components with audit data:

typography_avg = mean(comp.scores.typography for comp in components if score exists)
color_avg = mean(comp.scores.color for comp in components if score exists)
layout_avg = mean(comp.scores.layout for comp in components if score exists)
botanical_avg = mean(comp.scores.[DEPRECATED_STYLE] for comp in components if score exists)

Migration Compliance

Percentage of components using kerala-rage naming (not M3*):

migration = count(not name.startswith("M3")) / total_components

Use Cases

1. Daily Status Check

python3 .claude/skills/compliance-dashboard/run.py

Get instant health snapshot. Identify which components need attention today.

2. CI/CD Integration

python3 .claude/skills/compliance-dashboard/run.py --json > compliance-report.json

Export JSON for automated dashboards, slack notifications, or PR comments.

3. Component Deep Dive

python3 .claude/skills/compliance-dashboard/run.py --component Lens --json

Analyze single component: what's missing? What audit scores exist? What needs to be done?

4. Compliance Thresholds

  • 🟢 Green (≥80%): Production-ready
  • 🟡 Yellow (60-79%): Approaching threshold
  • 🔴 Red (<60%): Significant work needed

Error Handling

| Error | Behavior | |-------|----------| | Missing component-batch-plan.yaml | Exit code 1 | | --component <Name> not found | Exit code 1 | | Missing audit JSON | Scores = null, no error | | Missing component code/tests/specs | Flags = false, status = needs_refinement | | Inventory script fails | Continue with empty inventory data |

Integration with Other Skills

With kerala-rage-visual-audit

After running kerala-rage-visual-audit on components, the tool automatically picks up audit JSON files from docs/design/audits/ and includes scores in output.

With simple-component-inventory.sh

Tool runs this script automatically to detect:

  • Hardcoded colors/spacing/shadows
  • Duplicate component names

With component-batch-plan.yaml

Uses target state to determine:

  • Whether tests are required
  • Whether stage 2 is complete
  • Token strictness level
  • Component priority

Example: Full Workflow

# 1. Run compliance dashboard
python3 .claude/skills/compliance-dashboard/run.py

# Output shows Lens needs refinement with needs_typography_fix

# 2. Run visual audit on Lens
python3 .claude/skills/kerala-rage-visual-audit/run.py --component Lens

# 3. Review audit results, create docs/design/audits/Lens.json

# 4. Re-run compliance dashboard
python3 .claude/skills/compliance-dashboard/run.py

# Output now includes Lens audit scores + status update

Implementation

  • Language: Python 3
  • Location: .claude/skills/compliance-dashboard/run.py
  • Dependencies: PyYAML, standard library
  • Runtime: < 5 seconds (all components)

Exit Codes

  • 0: Success (even if compliance is low)
  • 1: Error (missing files, invalid input)

Limitations

✅ Tracks compliance across all components ✅ Integrates audit data automatically ✅ Detects hardcoded values and duplicates ✅ Validates against target state

❌ Cannot fix components (only identifies gaps) ❌ Requires manual audit JSON creation (can use kerala-rage-visual-audit skill) ❌ Metrics are relative to your standards, not benchmarks

Next Steps

After running the dashboard:

  1. Review top-priority needs (high priority + needs_refinement)
  2. Run visual audits on components missing scores
  3. Create audit JSON files (docs/design/audits/<Name>.json)
  4. Re-run dashboard to track progress
  5. Celebrate milestones (80% compliance reached!)

Design systems are measured by their ability to guide consistent, excellent work. This dashboard shows whether you're achieving that.