Skip to content
CryngePublic

About

Marketing AI

Topics

Resources

Stars

1 star

Watchers

0 watching

Forks

Latest commit

 

History

4 Commits

Folders and files

Repository files navigation

🎨 BrandMuse

AI brand voice management & content intelligence — ensure every piece of content speaks in your brand's voice with tone analysis, consistency auditing, and on-brand content generation across all channels.

CI TypeScript Python License Stars Last Commit

Voice Profile • Quick Start • Architecture • API • Modules • Contributing


⭐ Building a consistent brand? Star BrandMuse to support brand voice AI!


🎯 Brand Voice Profile

◤━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━◥
│                     BRAND VOICE PROFILE                        │
├━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┤
│                                                                │
│  Brand:          Acme Corp                                     │
│  Tone:           Professional · Approachable · Innovative      │
│  Vocabulary:     Industry terms · Avoid jargon                 │
│  Formality:      7/10   (formal but not stiff)                 │
│  Emotion:        Optimistic · Authoritative                    │
│  Audience:       CTOs and engineering leaders                  │
│  Channels:       Blog · LinkedIn · Email · Twitter             │
│                                                                │
│  Consistency:    ████████░░  82%   ⚠️ Needs improvement        │
│  Last Audit:     2026-07-01  ·  1,247 pieces analyzed          │
│  Top Violation:  Overuse of passive voice in blog posts        │
│                                                                │
◣━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━◢

Features

Feature Description Accuracy
Voice Analysis Scores tone, formality, emotion, vocabulary against brand guidelines 94% agreement with human raters
Content Generation Writes blog posts, social copy, and emails in your brand voice 4.2/5 quality score
Consistency Audit Scans existing content library for voice deviations Scans 1000 docs/min
Multi-language Maintains brand voice across English, Spanish, French, German Native-level fluency
API-first Integrate with CMS, email tools, and social schedulers REST + Webhook

Quick Start

npm install @crynge/brandmuse

# Analyze content against a brand profile
npx brandmuse analyze --content "Check out our latest product!" --profile acme

# Generate on-brand copy
npx brandmuse generate --brief "Launch email for v2.0" --tone professional

# Run a full brand audit
npx brandmuse audit --dir ./content --profile acme --format detailed
import { VoiceAnalyzer } from '@crynge/brandmuse/core/voice';

const analyzer = new VoiceAnalyzer({
  profile: {
    tone: ['professional', 'approachable'],
    formality: 0.7,
    vocabulary: ['ROI', 'scalable', 'enterprise'],
    disallowed: ['synergy', 'game-changer', 'utilize'],
  },
});

const result = await analyzer.analyze(
  "Our platform helps enterprises scale their infrastructure."
);

console.log({
  toneScore: result.toneScore,          // 0.88
  formalityFit: result.formalityFit,    // 0.92
  vocabularyMatch: result.vocabScore,   // 0.85
  violations: result.violations,        // []
  overallFit: result.overall,           // 0.88
});

Architecture

flowchart TB
    subgraph Input["Content Sources"]
        B[Blog Posts] --> I[Ingestion Pipeline]
        S[Social Media] --> I
        E[Email Campaigns] --> I
        W[Website Copy] --> I
    end

    subgraph Analysis["Analysis Layer"]
        I --> TA[Tone Analyzer]
        I --> CA[Consistency Auditor]
        I --> VA[Vocabulary Checker]

        TA --> TS[Tone Score]
        CA --> CS[Consistency Score]
        VA --> VS[Vocabulary Score]
    end

    subgraph Profile["Brand Profile Engine"]
        TS --> BP[Brand Profile]
        CS --> BP
        VS --> BP

        BP --> F1[Fit Score: 0.88]
        BP --> F2[Violations: 12]
        BP --> F3[Recommendations]
    end

    subgraph Generation["Generation Layer"]
        F3 --> GEN[Content Generator]
        GEN --> OUT[On-Brand Content]
    end

    subgraph API["API Layer"]
        BP --> API[REST API]
        GEN --> API
    end
Loading

API

# Analyze content
curl -X POST http://localhost:3000/api/analyze \
  -H "Content-Type: application/json" \
  -d '{"content": "Check out our product!", "profile": "acme"}'

# Generate content
curl -X POST http://localhost:3000/api/generate \
  -d '{"brief": "Newsletter intro", "profile": "acme", "tone": "professional"}'
# Python analyzer for batch processing
from brandmuse.analyzers.tone import ToneAnalyzer

analyzer = ToneAnalyzer()
scores = analyzer.analyze("We're revolutionizing the industry with AI.")
print(scores.formality)    # 0.45
print(scores.enthusiasm)   # 0.82

Modules

src/
├── core/
│   ├── voice.ts           # Voice profile matching engine
│   ├── auditor.ts         # Bulk content auditing
│   └── generator.ts       # On-brand content generation
├── analyzers/
│   ├── tone.py            # Tone classification (Python)
│   └── consistency.py     # Cross-document consistency checks
└── api/
    └── server.ts          # REST API server

Contributing

See CONTRIBUTING.md for guidelines.


License

MIT


🌐 Crynge Ecosystem

All repos are free and open-source. ⭐ Star what you use!

Category Repos
LLM & AI SpecInferKit · AetherAgents · PromptShield
Marketing AdVerify · Attributor · InfluencerHub · EdgePersona · AdVantage · BrandMuse · CampaignForge
Simulation CivSim · EvalScope
Operations OpsFlow
Built by Crynge · ⭐ Star us on GitHub!

About

Marketing AI

Topics

Resources

Stars

1 star

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages