epieczko-betty

Betty Framework

---

⚙️ Integration Note: Betty as a Claude Code Plugin System

Betty is designed as a collection of Claude Code plugins. You do not invoke Betty via a standalone CLI (betty command). Instead:

• Claude Code serves as the execution environment and unified CLI
• Each skill, agent, and command is registered through its manifest (skill.yaml, agent.yaml, etc.)
• Components automatically become discoverable and executable inside Claude Code's command palette
• All invocation, routing, and argument parsing is handled by Claude Code

No separate installation step is needed beyond plugin registration in your Claude Code environment.

---

🎯 What is Betty?

Betty Framework is RiskExec's AI-native SDLC platform built on Anthropic's Claude Code. Where Claude Code provides the runtime, Betty adds methodology, orchestration, and governance—transforming AI capabilities into a repeatable, auditable engineering discipline.

Named for Betty Shannon (1922-2017), the electrical engineer who translated Claude Shannon's information theory into applied computation, Betty embodies turning reasoning into engineered systems.

At a Glance

• 35 Production Skills — Composable building blocks for automation
• 20 Intelligent Agents — Orchestrators combining skills into workflows
• 6 Meta-Agents — AI-powered generators that create skills, agents, and hooks
• 391 Artifact Types — Enterprise SDLC standards (TOGAF, SAFe, Big Five)
• 406 Ready Templates — Consulting-quality artifacts across all phases
• Full Governance — Audit trails, policy enforcement, telemetry
• v1.0.0 Released — Production-ready, marketplace-certified

---

📑 Table of Contents

• Key Features
• Quick Start
• Meta-Agents: Self-Improving Infrastructure
• Artifact Framework
• Skills Catalog
• Agents Catalog
• Enterprise Features
• Documentation Guide
• Repository Structure
• Requirements
• Contributing
• License

---

✨ Key Features

🤖 Meta-Agents: Self-Improving Infrastructure

Betty includes 6 meta-agents that create and manage framework components autonomously:

Meta-Agent	Creates	Use Case
meta.agent	Agents from descriptions	"Create an API designer agent"
meta.skill	Skills from descriptions	"Generate a validation skill"
meta.hook	Event-driven hooks	"Add pre-commit API validation"
meta.artifact	Artifact type definitions	"Define deployment manifests"
meta.compatibility	Compatibility graphs	"Analyze agent dependencies"
meta.suggest	Next-step recommendations	"What should I build next?"

Breakthrough: Meta-agents enable recursive AI infrastructure where Betty builds Betty.

📦 Artifact Framework: 391 Enterprise Standards

Complete SDLC artifact library covering:

• Architecture (60+ types) — TOGAF, C4, ADRs, system designs
• Security (80+ types) — Threat models, pen tests, compliance
• Governance (100+ types) — RAID logs, charters, decision logs
• Data (40+ types) — Data models, catalogs, lineage
• Testing (50+ types) — Test plans, automation, performance
• Requirements (30+ types) — User stories, epics, BRDs
• Deployment (20+ types) — Runbooks, DR plans, SRE docs
• Plus: AI/ML, Operations, Compliance, Documentation

406 ready-to-use templates with Big Five consulting-quality content.

🛠️ 35 Production Skills

Composable, auditable building blocks across:

• Foundation — skill.create, registry.update, workflow.compose
• API Development — api.define, api.validate, api.generate-models
• Governance — policy.enforce, audit.log, telemetry.capture
• Infrastructure — plugin.build, plugin.sync, agent.run
• Documentation — docs.sync.readme, generate.docs