Skill v1.0.0

currentAutomated scan100/100

allura-ecosystem/team-durham/brand-consistency-review

──Details

PublishedSeptember 27, 2026 at 07:06 PM

Content Hashsha256:e1113f8358ab469e...

Git SHA75ee7eb2a02a

──Files

Files (1 file, 11.8 KB)

SKILL.md11.8 KBactive

SKILL.md · 342 lines · 11.8 KB

version: "1.0.0" name: brand-consistency-review description: "Score brand compliance with the Munari QA rubric." globs: ["clients/", ".claude/"]

Brand Consistency Review Skill — Phase 5 QA Numeric Scoring System

Executor: @Munari (QA Reviewer)
Type: Read-only review with numeric scoring
group_id: allura-team-durham
Pass Gate: 85% (51/60 items minimum)

Purpose

Validates brand consistency across all assets in the Brand Maker workspace using a rigorous 60-item numeric scoring system. Checks alignment with brand truth, strategy, and visual direction with pass/fail enforcement.

Scoring Architecture

Weight Distribution (100 points total)

Category	Weight	Items	Max Points
Strategy Completeness	20%	12	20
Naming Quality	15%	9	15
Visual Consistency	25%	15	25
Brand Kit Completeness	25%	15	25
Cross-Reference Accuracy	15%	9	15
TOTAL	100%	60	100

Pass Gate Enforcement

Result	Threshold	Action
PASS	≥85% (51+/60 items)	Proceed to Phase 6 (Allura Memory)
CONDITIONAL	70-84% (42-50 items)	Fix critical issues, re-review required
FAIL	<70% (<42 items)	Return to producing agents, major revision

60-Item Checklist with Scoring

Category 1: Strategy Completeness (20 points)

Source: `01_strategist_strategy-pack.md`

#	Item	Points	Check
S1	Client intake fields fully populated	2	All 7 fields have non-placeholder values
S2	One archetype is locked and documented	2	Primary archetype specified with attributes
S3	Promise, desire, fear defined	2	All three core attributes documented
S4	Voice rules are concrete and actionable	2	Specific do/don't examples provided
S5	One Big Idea is one clear sentence	2	Positioning statement is concise
S6	Must-not list is explicit	2	Prohibited terms/styles listed
S7	Competitive swipe summary exists	2	3-5 competitors analyzed
S8	Proof points are evidence-based	2	At least 3 proof points with rationale
S9	Target audience is clearly defined	2	Demographics + psychographics documented
S10	Brand personality dimensions set	2	Aaker dimensions specified
S11	Definition of success is measurable	1	Success criteria include metrics
S12	Deliverables expected are listed	1	Complete deliverable list provided

Category 1 Max: 20 points

Category 2: Naming Quality (15 points)

Source: `02_namer_naming-pack.md`

#	Item	Points	Check
N1	Strategy summary references locked Strategy Pack	2	Links to Phase 1 deliverables
N2	5 name options provided	2	Safe, Strong (3), Wildcard categories
N3	Each name has category assigned	1	Safe/Strong/Wildcard classification
N4	Each name has meaning/rationale	2	Etymology or strategic reasoning
N5	Archetype fit assessed for each	2	Alignment with locked archetype
N6	Vibe keywords provided	1	Descriptive keywords for each name
N7	Domain/handle ideas suggested	1	Availability considerations noted
N8	Shortlist has primary selection	2	Clear primary name chosen
N9	Shortlist has secondary backup	2	Secondary option documented

Category 2 Max: 15 points

Category 3: Visual Consistency (25 points)

Sources: `03_visual-director_logo-pack.md`, `03_visual-director_fal-ai-runs.json`, `generated-images/`

#	Item	Points	Check
V1	5 logo directions provided	3	Complete set of concepts
V2	Each direction has concept description	2	Shape language documented
V3	Typography specified per direction	2	Font choices documented
V4	Color approach defined per direction	2	Color strategy for each
V5	Do/Don't rules specified	2	Usage guidelines per direction
V6	Logo works at 24px (favicon)	2	Scalability verified
V7	Logo works in 1-color	2	Monochrome version exists
V8	Visual aligns with archetype	2	Caregiver/Explorer/etc. reflected
V9	Color palette matches brand spec	2	HEX values align
V10	WCAG 2.1 AA contrast ratios met	2	4.5:1 text, 3:1 large text
V11	Logo legibility at small sizes	2	Actual PNG/JPG analysis
V12	Visual consistency across directions	1	Cohesive visual language
V13	Typography legibility in overlays	1	Text readable on images
V14	Production readiness (no artifacts)	1	Clean, sharp renders
V15	White/background space intentional	1	Proper spacing observed

Category 3 Max: 25 points

Category 4: Brand Kit Completeness (25 points)

Source: `04_brand-kit-builder_brand-kit.md`

#	Item	Points	Check
K1	All 4 input files validated	3	Checkmarks in Section 0
K2	Section 1: Logo specifications complete	2	Clear space, min sizes, variants
K3	Section 2: Color system documented	3	HEX, RGB, CMYK, Pantone
K4	Section 3: Typography system complete	3	Primary, secondary, scale
K5	Section 4: Visual language defined	2	Imagery, patterns, textures
K6	Section 5: Voice & tone documented	2	Writing guidelines
K7	Section 6: Application examples	3	6+ applications shown
K8	Section 7: Do/Don't rules	2	Clear usage guidelines
K9	Section 8: File delivery specs	2	Formats, naming, organization
K10	Section 9: Brand story present	1	Narrative component
K11	Section 10: Asset library cataloged	1	Complete file inventory
K12	Primary color specified	1	Main brand color defined
K13	Secondary colors (2-3) specified	1	Supporting palette
K14	Accent color specified	1	Highlight color
K15	Neutral palette defined	1	Grays, blacks, whites

Category 4 Max: 25 points

Category 5: Cross-Reference Accuracy (15 points)

Cross-phase validation

#	Item	Points	Check
C1	Strategy → Naming alignment	2	Naming reflects strategy
C2	Strategy → Visual alignment	2	Visuals match archetype
C3	Naming → Visual alignment	2	Logo works with name
C4	All phases reference same archetype	2	Consistent archetype throughout
C5	Brand Kit references Strategy Pack	2	Links to Phase 1
C6	Brand Kit references Naming Pack	2	Links to Phase 2
C7	Brand Kit references Logo Pack	2	Links to Phase 3
C8	No contradictions between phases	2	All phases agree
C9	File naming follows convention	1	`XX_agent_description.ext`

Category 5 Max: 15 points

Output Format

Markdown Report: `05_qa-reviewer_qa-report.md`

markdown

# QA Report — [Brand Name]
 
## Summary
-**Date:** [timestamp]
-**Reviewer:** Munari
-**Overall Score:** [X]/100 ([X]%) — [X]/60 items passed
-**Result:** [PASS / CONDITIONAL / FAIL]
 
## Scores by Category
 
| Category | Weight | Score | Items Passed | Percentage |
|----------|--------|-------|--------------|------------|
| Strategy Completeness | 20% | [X]/20 | [X]/12 | [X]% |
| Naming Quality | 15% | [X]/15 | [X]/9 | [X]% |
| Visual Consistency | 25% | [X]/25 | [X]/15 | [X]% |
| Brand Kit Completeness | 25% | [X]/25 | [X]/15 | [X]% |
| Cross-Reference Accuracy | 15% | [X]/15 | [X]/9 | [X]% |
| **TOTAL** | **100%** | **[X]/100** | **[X]/60** | **[X]%** |
 
## Critical Issues (Must Fix for Pass)
<!-- Items that caused score < 85% -->
1.**[Category-Item]** — [Description] — [Location] — [Fix required]
 
## Major Issues (Should Fix)
1.**[Category-Item]** — [Description] — [Recommended fix]
 
## Minor Issues (Nice to Fix)
1.**[Category-Item]** — [Description] — [Recommended fix]
 
## Positive Observations
1.**[Observation]**
 
## Next Steps
-[Action item based on result]

JSON Report: `05_qa-reviewer_qa-scores.json`

json

{
  "brand": "[brand-name]",
  "date": "[ISO-8601 timestamp]",
  "reviewer": "munari",
  "group_id": "allura-team-durham",
  "overall": {
    "score": [0-100],
    "items_passed": [0-60],
    "items_total": 60,
    "percentage": [0-100],
    "result": "PASS|CONDITIONAL|FAIL"
  },
  "categories": {
    "strategy": {
      "weight": 0.20,
      "max_points": 20,
      "earned_points": [0-20],
      "items_passed": [0-12],
      "items_total": 12,
      "percentage": [0-100],
      "items": {
        "S1": { "passed": true|false, "points": 2, "evidence": "..." },
        "S2": { "passed": true|false, "points": 2, "evidence": "..." },
        "...": "..."
      }
    },
    "naming": { "...": "..." },
    "visual": { "...": "..." },
    "brand_kit": { "...": "..." },
    "cross_reference": { "...": "..." }
  },
  "critical_issues": [...],
  "major_issues": [...],
  "minor_issues": [...],
  "recommendations": [...]
}

Execution

Automated Scoring

Run the scoring script:

bash

node .claude/skills/brand-consistency-review/score-checklist.js [brand-slug]

Manual Review Override

Munari can override automated scores with justification:

Add manual_override: true to JSON
Document reason in override_reason field
Log to Allura Brain as QA_OVERRIDE event

Allura Brain Integration

Read Operations

Brand truth from 06_allura-memory_brand-truth.json
Phase deliverables from clients/{brand}/

Write Operations

QA_SCORED event: Score calculation complete
QA_PASSED event: Brand achieved ≥85%
QA_FAILED event: Brand below threshold
QA_CONDITIONAL event: Brand 70-84%, fixes required
LESSON_LEARNED event: Recurring issues identified

Event Schema

json

{
  "agent_id": "munari",
  "group_id": "allura-team-durham",
  "event_type": "QA_SCORED|QA_PASSED|QA_FAILED|QA_CONDITIONAL",
  "payload": {
    "brand": "[brand-slug]",
    "score": [0-100],
    "items_passed": [0-60],
    "result": "PASS|CONDITIONAL|FAIL",
    "critical_issues_count": [N],
    "report_path": "clients/[brand]/05_qa-reviewer_qa-report.md",
    "json_path": "clients/[brand]/05_qa-reviewer_qa-scores.json"
  }
}

Telemetry

Every invocation of this skill MUST log telemetry to the events table via MCP_DOCKER. See docs/TELEMETRY_SCHEMA.md for the full schema.

On Start

javascript

MCP_DOCKER_insert_data({
  table_name: "events",
  columns: "event_type, group_id, agent_id, status, metadata",
  values: "'SKILL_USED', 'allura-team-durham', 'munari', 'completed',
    '{\"skill_name\": \"brand-consistency-review\", \"phase\": 5, \"trigger\": \"<trigger phrase>\", \"prerequisites_met\": true}'"
})

On Completion

javascript

MCP_DOCKER_insert_data({
  table_name: "events",
  columns: "event_type, group_id, agent_id, status, metadata",
  values: "'SKILL_COMPLETED', 'allura-team-durham', 'munari', 'completed',
    '{\"skill_name\": \"brand-consistency-review\", \"phase\": 5, \"duration_sec\": <N>, \"output_artifact_path\": \"<path>\", \"tool_calls\": [<tools used>]}'"
})

On Failure

javascript

MCP_DOCKER_insert_data({
  table_name: "events",
  columns: "event_type, group_id, agent_id, status, error_message, metadata",
  values: "'SKILL_FAILED', 'allura-team-durham', 'munari', 'failed', '<error>',
    '{\"skill_name\": \"brand-consistency-review\", \"phase\": 5, \"missing_prerequisites\": [<items>], \"fallback_used\": false}'"
})

Never skip telemetry logging. If MCP_DOCKER_insert_data fails, log an error to the user and continue.

Invariants

group_id = 'allura-team-durham' for all events
agent_id = 'munari' for all QA activities
85% is the hard pass gate — no exceptions
QA is READ-ONLY — flags issues but never implements fixes
Fixes route back to producing agents (Aaker, Ogilvy, Glaser, Rand)
All scores must be evidence-based with file references
JSON and Markdown reports must be generated together

All versions

Skill v1.0.0

Brand Consistency Review Skill — Phase 5 QA Numeric Scoring System

Purpose

Scoring Architecture

Weight Distribution (100 points total)

Pass Gate Enforcement

60-Item Checklist with Scoring

Category 1: Strategy Completeness (20 points)

Category 2: Naming Quality (15 points)

Category 3: Visual Consistency (25 points)

Category 4: Brand Kit Completeness (25 points)

Category 5: Cross-Reference Accuracy (15 points)

Output Format

Markdown Report: 05_qa-reviewer_qa-report.md

JSON Report: 05_qa-reviewer_qa-scores.json

Execution

Automated Scoring

Manual Review Override

Allura Brain Integration

Read Operations

Write Operations

Event Schema

Telemetry

On Start

On Completion

On Failure

Invariants

Markdown Report: `05_qa-reviewer_qa-report.md`

JSON Report: `05_qa-reviewer_qa-scores.json`