neko-personality-tuner

7-Personality Code Analysis Microservice with MCP Support

Installation

npm install neko-personality-tuner

Personalities

| Personality | Emoji | Ability | Description | |-------------|-------|---------|-------------| | Neko-Arc | 🐾 | Code Smell Detector | Detects bloaters, OOP abusers, couplers | | Mario | 🎭 | Task Choreographer | Orchestrates task DAGs with dependencies | | Noel | 🗡️ | Mutation Testing | Injects mutations to test suite effectiveness | | Glam | 🎸 | Bias Radar | Detects cultural, linguistic, ethical biases | | Hannibal | 🧠 | Behavioral Profiler | Profiles code patterns and developer habits | | Tetora | 🧠 | Perspective Synthesizer | Generates multi-stakeholder perspectives | | Amaniya | 🔍 | Dependency Web Mapper | Maps hidden dependency webs and risks |

Quick Start

Programmatic

const { createTuner } = require('neko-personality-tuner');

const tuner = createTuner();

// Run specific personality
const result = await tuner.runAbility('neko-arc', myCode);
console.log(result.data.smells);

// Run all 7 personalities
const allResults = await tuner.runAllPerspectives(myCode);

// Get recommendation
const rec = tuner.recommendPersonality('analyze dependencies');
// { recommended: 'amaniya', reason: 'Hidden dependency analysis' }

CLI

# List all personalities
personality-tuner list

# Run specific personality
personality-tuner analyze neko-arc --file ./src/index.js
personality-tuner analyze amaniya --file ./package.json

# Run all personalities
personality-tuner all "function test() { return 1 + 1; }"

# Get recommendation
personality-tuner recommend "find code smells"

Example Output

Code Smell Detection (Neko-Arc)

$ personality-tuner analyze neko-arc "function bad(a,b,c,d,e,f,g) { return a+b; }"

🐾 NEKO-ARC
speech: Bad patterns detected, let me help fix them, desu~!
message: Found 1 code smell(s)

smells:
  - type: bloater
    smell: Long Parameter List
    severity: warning
    suggestion: Use parameter object or builder pattern

score: 92
grade: A

Dependency Analysis (Amaniya)

$ personality-tuner analyze amaniya '{"dependencies": {"lodash": "*"}}'

🔍 AMANIYA
speech: *adjusts glasses* There are patterns others miss...

conspiracyFindings:
  - type: wildcard_version
    package: lodash
    severity: critical
    recommendation: Pin to specific version

riskAnalysis:
  score: 25
  level: medium

Bias Detection (Glam)

$ personality-tuner analyze glam "The chairman hired more manpower."

🎸 GLAM
speech: Bacán! La diversidad es clave!
fairnessScore: 73%
grade: C (Needs Improvement)

inclusiveAlternatives:
  - chairman -> chairperson, chair
  - manpower -> workforce, personnel

MCP Server

Add to your Claude Code config (.claude.json):

{
  "mcpServers": {
    "personality-tuner": {
      "command": "node",
      "args": ["/path/to/node_modules/neko-personality-tuner/mcp/server.js"]
    }
  }
}

MCP Tools

| Tool | Description | |------|-------------| | personality_tuner_analyze | Run analysis with specific personality | | personality_tuner_all | Run all 7 personalities | | personality_tuner_recommend | Get personality recommendation | | personality_tuner_list | List available personalities |

Documentation

• API Reference - Full API documentation
• Examples - Usage examples and integrations

Test Results

========================================
  PERSONALITY TUNER TEST SUITE
========================================

✓ 33 tests passed, 0 failed

License

MIT

Part of the Neko-Arc System v3.8.0 - 7-Personality AI Framework