Professor Hudak plugin

Overview

Professor Hudak is a discovery assistant for the curated AI assets in the otc-awesome-llm repository. It helps you navigate prompts, chatmodes, instructions, collections, and skills with task-aware recommendations. Asset maturity varies — the repo is also the staging ground for products graduating to standalone repositories (Dr Zero, claude-automation-plugins, Nightingale), so treat the assets as a working library rather than a stable distribution.

The plugin provides five slash commands for asset discovery, recent updates, contribution guidance, task-based recommendations, and general help. It connects to the otc-awesome-llm MCP server for real-time asset data, analyzes git history for recent changes, and integrates with quality scoring for asset rankings.

Version: 7.4.1 Author: Thomas Hudak

Installation

claude plugin install professor-hudak@otc-awesome-llm

Prerequisites

• Claude Code CLI
• Node.js 18 or later
• Access to the otc-awesome-llm repository

After installation, verify:

claude plugin list
# Should show "professor-hudak" in the output

Commands

/ph-discover -- Search and discover AI assets

Search across all asset types by keyword, tag, or domain.

/ph-discover kubernetes
/ph-discover terraform review
/ph-discover chatmode
/ph-discover tag:ops

Results are grouped by asset type (prompts, chatmodes, instructions, collections) and include quality scores when available.

Example output:

Found 8 Kubernetes-related assets:

Prompts (3):
  k8s-pod-debug         - Debug failing pods (Quality: 92%)
  k8s-deployment-rollout - Safe deployment strategies
  k8s-resource-optimization - Resource limit tuning

Chatmodes (1):
  k8s-operations-assistant - Long-running K8s helper

Collections (1):
  ops-k8s-core - Complete K8s operations bundle

/ph-recommend -- Get task-based recommendations

Describe what you are working on and receive ranked recommendations across all asset types.

/ph-recommend debug kubernetes pod crash
/ph-recommend plan terraform azure infrastructure
/ph-recommend write ansible playbook for nginx
/ph-recommend improve code review process

Recommendations are ranked by:

• Direct keyword matches
• Tag alignment
• Domain relevance
• Task type alignment (debug, create, review, optimize)
• Quality scores

The output is organized into categories:

• Perfect match -- Assets directly targeting your task (prompts).
• For extended work -- Long-running sessions for the domain (chatmodes).
• Complete bundle -- End-to-end workflow packages (collections).
• Apply these standards -- Coding standards and safety rules (instructions).

When appropriate, Professor Hudak also suggests Dr. Zero modes (autonomous, refinement, execution, or analysis) for the task.

/ph-contribute -- Get contribution guidance

Get step-by-step guidance for contributing a new asset to otc-awesome-llm.

/ph-contribute              # Interactive mode (asks what type)
/ph-contribute prompt       # Template for a new prompt
/ph-contribute chatmode     # Template for a new chatmode
/ph-contribute instruction  # Template for a new instruction

The command provides:

1. A complete YAML frontmatter template for the chosen asset type.
2. File path guidance (shared/<type>s/your-name.<type>.md).
3. A validation checklist (npm run validate, npm test, npm run quality-check).
4. Conventional commit message format.

/ph-help -- Get help and overview

Get general help or topic-specific guidance.

/ph-help                    # General overview and quick stats
/ph-help getting-started    # First steps with otc-awesome-llm
/ph-help asset-types        # Understanding prompts, chatmodes, instructions, etc.
/ph-help dr-zero            # Learn about autonomous improvement modes

Quick stats (as of the latest index):

/ph-recent -- Show recent updates

Display recent additions and changes to the repository.

/ph-recent          # Last 7 days (default)
/ph-recent 14       # Last 14 days
/ph-recent 30       # Last month

Output categorizes changes into new assets, updated assets, releases, and Dr. Zero improvements.

Architecture

Professor Hudak consists of three layers:

Commands (/ph-*)
    |
    v
Agents (professor-hudak-core)
    |
    v
Source modules (src/)
    asset-indexer.js          -- Indexes and searches repository assets
    recommendation-engine.js  -- Provides intelligent suggestions
    update-tracker.js         -- Tracks git history and changes
    utils/exec-file-no-throw.js -- Safe command execution (no shell injection)

Integration points

Security

• Uses execFile instead of exec to prevent command injection.
• All user inputs are sanitized before use.
• Read-only access to repository data.
• No file modifications without explicit user action.

Usage example

A typical workflow combining multiple Professor Hudak commands:

# 1. Find out what is new this week
/ph-recent

# 2. Discover assets related to your current domain
/ph-discover terraform azure

# 3. Get recommendations for your specific task
/ph-recommend plan terraform module for azure key vault

# 4. Start contributing a new asset based on a gap you found
/ph-contribute prompt

Troubleshooting

Related

• Plugin catalog -- Overview of all available plugins
• otc-awesome-llm repository -- Source repository containing all assets and plugins