CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Repository Overview

This is the Public Radio Agents project - a specialized AI framework that transforms Large Language Models into expert consultants for public radio station management. The project consists of two main components:

Core Framework - A BMAd-Method™ adaptation that creates 4 specialized agents for public radio management
SaaS Frontend - A React/Tailwind web application for hosted access to the framework

Key Architecture

Core Framework Structure

Main Bundle:
```
publicradio.txt
```
(21KB) - Complete framework ready for any LLM
Individual Agents: 4 specialized directors in
```
agents/
```
directory
Dependencies: 113+ knowledge bases, templates, and checklists in
```
agents/dependencies/
```
Scripts: Python tools for validation and bundle building

SaaS Frontend Structure

Framework: Next.js 14 with TypeScript and Tailwind CSS
Components: React components for agent interaction, chat interface, and station management
Target Deployment: Vercel with Supabase database integration

Essential Commands

Framework Development Commands

# Validate framework integrity and dependencies
python3 scripts/validate-dependencies.py

# Build complete framework bundle from components
python3 scripts/build-bundle.py

# Build custom bundle to specific location
python3 scripts/build-bundle.py --output custom-bundle.txt

SaaS Frontend Commands

# Development server (from saas-frontend directory)
cd saas-frontend && npm run dev

# Build for production
npm run build

# Type checking
npm run type-check

# Linting
npm run lint

Project Architecture

Public Radio Framework Architecture

The framework operates on a command-driven interface where all commands start with

Orchestrator Agent (
```
bmad-orchestrator
```
) - Routes requests and manages multi-agent workflows
Development Director - Fundraising, membership campaigns, donor relations
Marketing Director - Audience development, digital marketing, community engagement
Underwriting Director - Corporate partnerships, sponsorship sales, business development
Program Director - Programming strategy, content development, FCC compliance

Agent Dependencies Structure

Each agent has 4 dependency types:

data/
- Domain knowledge bases (.md files)
tasks/
- Specific methodologies (.md files)
templates/
- YAML document templates (.yaml files)
checklists/
- Quality assurance checklists (.md files)

SaaS Frontend Architecture

src/
├── components/
│   ├── agents/          # Agent selection and switching UI
│   ├── chat/            # Chat interface and message handling
│   └── station/         # Station profile management
├── types/               # TypeScript type definitions
└── lib/                 # Utilities and integrations

Key TypeScript Types

The SaaS frontend uses comprehensive TypeScript interfaces:

Station
- Public radio station profiles with size, license type, budget, challenges
Agent
- Agent definitions with personas, expertise, and available commands
ChatMessage
- Chat system with role-based messaging and metadata
Workflow
- Multi-phase structured processes with deliverables

Framework Usage Patterns

Command Structure

All framework interactions use

prefix commands:

*agent [name]

- Switch to specific agent (e.g.,

*agent development-director

)

*workflow [name]

- Start structured workflow (e.g.,

*workflow membership-campaign

)

```
*help
```
- Show available commands and agents
```
*chat-mode
```
- Begin conversational assistance
```
*kb-mode
```
- Access knowledge base

Agent Specializations

Development Director (
```
development-director
```
) - Expertise in fundraising strategies, donor psychology, grant writing, membership campaigns
Marketing Director (
```
marketing-director
```
) - Specializes in audience research, digital marketing, community engagement, brand management
Underwriting Director (
```
underwriting-director
```
) - Focuses on corporate partnerships, FCC compliance for underwriting, sponsorship packages
Program Director (
```
program-director
```
) - Handles programming strategy, content development, talent management, regulatory compliance

Validation and Quality Assurance

The framework includes automated validation:

Dependency Validation - Ensures all referenced files exist and are properly structured
YAML Validation - Verifies template syntax and required fields
Content Quality - Checks for appropriate length and markdown formatting
Orphaned File Detection - Identifies unreferenced dependency files

Development Best Practices

Working with Framework Components

Always run
```
validate-dependencies.py
```
before committing changes to ensure framework integrity
Use
```
build-bundle.py
```
to regenerate
```
publicradio.txt
```
after modifying individual agent components
Follow existing YAML template patterns when creating new templates
Maintain consistent markdown formatting in knowledge base files

SaaS Frontend Development

Follow existing component patterns in
```
components/agents/
```
and
```
components/chat/
```
Use the comprehensive TypeScript types defined in
```
src/types/index.ts
```
Implement proper error boundaries and loading states for chat interactions
Maintain responsive design principles with Tailwind CSS classes

File Organization

Individual agent configurations are in
```
agents/[agent-name]_agent.md
```
All dependencies are organized by agent in
```
agents/dependencies/[agent-name]/[type]/
```
Documentation follows a structured approach in the
```
docs/
```
directory
The main bundle
```
publicradio.txt
```
is generated automatically - do not edit manually

Testing and Validation

Framework Testing

The framework includes comprehensive validation scripts that check:

YAML syntax in agent configurations and templates
Markdown formatting in knowledge bases and tasks
File existence for all dependencies referenced in agent configurations
Content quality (minimum length, proper headers)

Common Issues

Missing dependency files result in validation errors
Incorrect YAML formatting breaks agent loading
Orphaned files indicate unused dependencies that should be removed or referenced

Integration Points

LLM Integration

The framework is designed to work with any chat-based LLM:

Load entire
```
publicradio.txt
```
as system context
Parse commands starting with
```
*
```
for agent switching and workflows
Maintain station context throughout conversations

SaaS Platform Integration

Chat interface parses framework commands and routes to appropriate agents
Station profiles provide context for personalized responses
Export functionality allows saving conversations and generated documents
Analytics track usage patterns across agents and workflows

Documentation Structure

Key documentation files:

```
docs/quick-start.md
```
- Getting started guide for new users
```
docs/user-guide.md
```
- Comprehensive usage documentation
```
docs/saas-frontend-guide.md
```
- Complete SaaS implementation guide
```
docs/ide-integration.md
```
- Integration with development environments
```
docs/case-study.md
```
- Real-world usage example

This repository represents a complete ecosystem for public radio station management using AI, from the core framework to a production-ready SaaS application.

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Repository Overview

Core Framework - A BMAd-Method™ adaptation that creates 4 specialized agents for public radio management
SaaS Frontend - A React/Tailwind web application for hosted access to the framework

Key Architecture

Core Framework Structure

Main Bundle:
```
publicradio.txt
```
(21KB) - Complete framework ready for any LLM
Individual Agents: 4 specialized directors in
```
agents/
```
directory
Dependencies: 113+ knowledge bases, templates, and checklists in
```
agents/dependencies/
```
Scripts: Python tools for validation and bundle building

SaaS Frontend Structure

Framework: Next.js 14 with TypeScript and Tailwind CSS
Components: React components for agent interaction, chat interface, and station management
Target Deployment: Vercel with Supabase database integration

Essential Commands

Framework Development Commands

# Validate framework integrity and dependencies
python3 scripts/validate-dependencies.py

# Build complete framework bundle from components
python3 scripts/build-bundle.py

# Build custom bundle to specific location
python3 scripts/build-bundle.py --output custom-bundle.txt

SaaS Frontend Commands

# Development server (from saas-frontend directory)
cd saas-frontend && npm run dev

# Build for production
npm run build

# Type checking
npm run type-check

# Linting
npm run lint

Project Architecture

Public Radio Framework Architecture

The framework operates on a command-driven interface where all commands start with

Orchestrator Agent (
```
bmad-orchestrator
```
) - Routes requests and manages multi-agent workflows
Development Director - Fundraising, membership campaigns, donor relations
Marketing Director - Audience development, digital marketing, community engagement
Underwriting Director - Corporate partnerships, sponsorship sales, business development
Program Director - Programming strategy, content development, FCC compliance

Agent Dependencies Structure

Each agent has 4 dependency types:

data/
- Domain knowledge bases (.md files)
tasks/
- Specific methodologies (.md files)
templates/
- YAML document templates (.yaml files)
checklists/
- Quality assurance checklists (.md files)

SaaS Frontend Architecture

src/
├── components/
│   ├── agents/          # Agent selection and switching UI
│   ├── chat/            # Chat interface and message handling
│   └── station/         # Station profile management
├── types/               # TypeScript type definitions
└── lib/                 # Utilities and integrations

Key TypeScript Types

The SaaS frontend uses comprehensive TypeScript interfaces:

Station
- Public radio station profiles with size, license type, budget, challenges
Agent
- Agent definitions with personas, expertise, and available commands
ChatMessage
- Chat system with role-based messaging and metadata
Workflow
- Multi-phase structured processes with deliverables

Framework Usage Patterns

Command Structure

All framework interactions use

prefix commands:

*agent [name]

- Switch to specific agent (e.g.,

*agent development-director

)

*workflow [name]

- Start structured workflow (e.g.,

*workflow membership-campaign

)

```
*help
```
- Show available commands and agents
```
*chat-mode
```
- Begin conversational assistance
```
*kb-mode
```
- Access knowledge base

Agent Specializations

Development Director (
```
development-director
```
) - Expertise in fundraising strategies, donor psychology, grant writing, membership campaigns
Marketing Director (
```
marketing-director
```
) - Specializes in audience research, digital marketing, community engagement, brand management
Underwriting Director (
```
underwriting-director
```
) - Focuses on corporate partnerships, FCC compliance for underwriting, sponsorship packages
Program Director (
```
program-director
```
) - Handles programming strategy, content development, talent management, regulatory compliance

Validation and Quality Assurance

The framework includes automated validation:

Dependency Validation - Ensures all referenced files exist and are properly structured
YAML Validation - Verifies template syntax and required fields
Content Quality - Checks for appropriate length and markdown formatting
Orphaned File Detection - Identifies unreferenced dependency files

Development Best Practices

Working with Framework Components

Always run
```
validate-dependencies.py
```
before committing changes to ensure framework integrity
Use
```
build-bundle.py
```
to regenerate
```
publicradio.txt
```
after modifying individual agent components
Follow existing YAML template patterns when creating new templates
Maintain consistent markdown formatting in knowledge base files

SaaS Frontend Development

Follow existing component patterns in
```
components/agents/
```
and
```
components/chat/
```
Use the comprehensive TypeScript types defined in
```
src/types/index.ts
```
Implement proper error boundaries and loading states for chat interactions
Maintain responsive design principles with Tailwind CSS classes

File Organization

Individual agent configurations are in
```
agents/[agent-name]_agent.md
```
All dependencies are organized by agent in
```
agents/dependencies/[agent-name]/[type]/
```
Documentation follows a structured approach in the
```
docs/
```
directory
The main bundle
```
publicradio.txt
```
is generated automatically - do not edit manually

Testing and Validation

Framework Testing

The framework includes comprehensive validation scripts that check:

YAML syntax in agent configurations and templates
Markdown formatting in knowledge bases and tasks
File existence for all dependencies referenced in agent configurations
Content quality (minimum length, proper headers)

Common Issues

Missing dependency files result in validation errors
Incorrect YAML formatting breaks agent loading
Orphaned files indicate unused dependencies that should be removed or referenced

Integration Points

LLM Integration

The framework is designed to work with any chat-based LLM:

Load entire
```
publicradio.txt
```
as system context
Parse commands starting with
```
*
```
for agent switching and workflows
Maintain station context throughout conversations

SaaS Platform Integration

Chat interface parses framework commands and routes to appropriate agents
Station profiles provide context for personalized responses
Export functionality allows saving conversations and generated documents
Analytics track usage patterns across agents and workflows

Documentation Structure

Key documentation files:

```
docs/quick-start.md
```
- Getting started guide for new users
```
docs/user-guide.md
```
- Comprehensive usage documentation
```
docs/saas-frontend-guide.md
```
- Complete SaaS implementation guide
```
docs/ide-integration.md
```
- Integration with development environments
```
docs/case-study.md
```
- Real-world usage example

This repository represents a complete ecosystem for public radio station management using AI, from the core framework to a production-ready SaaS application.

CLAUDE.md

Related Skills

Nano Banana Pro

Markdown Converter

1password