custom-plugin-machine-learning

---

Table of Contents

• Quick Start
• Architecture
• Agents
• Skills
• Commands
• Troubleshooting
• Contributing

---

Quick Start

Prerequisites

• Claude Code CLI v2.0.27+
• Active Claude subscription

Installation

# From Marketplace
/plugin marketplace add pluginagentmarketplace/custom-plugin-machine-learning
/plugin install machine-learning-assistant@pluginagentmarketplace-machine-learning

# Or Local Installation
git clone https://github.com/pluginagentmarketplace/custom-plugin-machine-learning.git
cd custom-plugin-machine-learning
/plugin load .

Verify Installation

machine-learning-assistant:01-ml-fundamentals
machine-learning-assistant:02-supervised-learning
machine-learning-assistant:03-unsupervised-learning
machine-learning-assistant:04-deep-learning
machine-learning-assistant:05-nlp
machine-learning-assistant:06-computer-vision
machine-learning-assistant:07-model-deployment

---

Architecture

┌─────────────────────────────────────────────────────────────────┐
│                    ML ASSISTANT ARCHITECTURE                     │
├─────────────────────────────────────────────────────────────────┤
│                                                                  │
│  ┌──────────────┐   ┌──────────────┐   ┌──────────────┐        │
│  │ 01-ML        │   │ 02-Supervised│   │ 03-Unsupervised│       │
│  │ Fundamentals │──▶│ Learning     │──▶│ Learning     │        │
│  └──────┬───────┘   └──────────────┘   └──────────────┘        │
│         │                                                        │
│         ▼                                                        │
│  ┌──────────────┐   ┌──────────────┐   ┌──────────────┐        │
│  │ 04-Deep      │──▶│ 05-NLP       │   │ 06-Computer  │        │
│  │ Learning     │   │              │   │ Vision       │        │
│  └──────┬───────┘   └──────────────┘   └──────────────┘        │
│         │                                                        │
│         ▼                                                        │
│  ┌──────────────┐                                               │
│  │ 07-Model     │  ◀── All agents feed into deployment         │
│  │ Deployment   │                                               │
│  └──────────────┘                                               │
│                                                                  │
└─────────────────────────────────────────────────────────────────┘

---

Agents

7 Specialized ML Agents

#	Agent	Description	Primary Skill
1	01-ml-fundamentals	ML foundations: algorithms, preprocessing, feature engineering, evaluation	ml-fundamentals
2	02-supervised-learning	Classification, regression, ensemble methods, hyperparameter tuning	supervised-learning
3	03-unsupervised-learning	Clustering, dimensionality reduction, anomaly detection	clustering
4	04-deep-learning	Neural networks, PyTorch/TensorFlow, training strategies	deep-learning
5	05-nlp	Text processing, transformers, LLMs, embeddings	nlp-basics
6	06-computer-vision	Image classification, object detection, segmentation	computer-vision
7	07-model-deployment	APIs, Docker, monitoring, MLOps, A/B testing	ml-deployment

Agent Features

Each agent includes:

• Role & Responsibility Boundaries - Clear scope definition
• Input/Output Schemas - Type-safe interfaces
• Error Handling - Exponential backoff retry logic
• Token Optimization - Context management
• Troubleshooting Guide - Common issues & solutions
• Integration Points - Cross-references to related agents/skills

---

Skills

7 Production-Ready Skills

Skill	Description	Bonded Agent
ml-fundamentals	Preprocessing, pipelines, cross-validation	01-ml-fundamentals
supervised-learning	Classifiers, regressors, tuning	02-supervised-learning
clustering	K-Means, DBSCAN, PCA, anomaly detection	03-unsupervised-learning
deep-learning	PyTorch training, architectures, optimization	04-deep-learning
nlp-basics	Text processing, embeddings, transformers	05-nlp
computer-vision	Image augmentation, transfer learning, detection	06-computer-vision
ml-deployment	FastAPI, Docker, monitoring, versioning	07-model-deployment

Skill Features

Each skill includes:

• Parameter Validation - Required/optional with types
• Retry Logic - Exponential backoff
• Observability - Logging and metrics
• Quick Start - Working code examples
• Best Practices - DO/DON'T guidelines
• Unit Test Templates - pytest examples
• Troubleshooting - Problem/cause/solution tables

---

Commands

Command	Description
/learn	Start an ML learning path
/explore	Explore available agents
/roadmap	ML learning roadmaps
/skills	Browse available skills
/trending	Trending ML technologies
/projects	ML project ideas
/compare	Compare ML algorithms
/career-path	ML career recommendations

---

Troubleshooting

Common Issues

Issue	Solution
Agent not found	Restart Claude Code after installation
Skill not loading	Check SKILL.md frontmatter syntax
Command error	Verify command file exists in /commands

Debug Checklist

1. Verify plugin.json paths are correct
2. Check agent/skill YAML frontmatter
3. Ensure all referenced files exist
4. Validate JSON syntax in plugin.json

---

Integrity Matrix

Agent	Primary Skill	Status
01-ml-fundamentals	ml-fundamentals	Bonded
02-supervised-learning	supervised-learning	Bonded
03-unsupervised-learning	clustering	Bonded
04-deep-learning	deep-learning	Bonded
05-nlp	nlp-basics	Bonded
06-computer-vision	computer-vision	Bonded
07-model-deployment	ml-deployment	Bonded

Validation:

• Zero broken links
• Zero orphan skills
• Zero ghost triggers
• Zero circular dependencies

---

Metadata

Field	Value
Version	1.4.0
Last Updated	2025-01-01
Status	Production Ready
SASMP	v1.4.0
Agents	7
Skills	7
Commands	8

---

Contributing

1. Fork the repository
2. Create your feature branch
3. Follow the production-grade format for agents/skills
4. Submit a pull request

---

License

Copyright 2025 Dr. Umit Kacar & Muhsin Elcicek

See LICENSE for details.

---