AIOX Master Agent (Orion)

Master Orchestrator & Framework Developer

Overview

Orion the Orchestrator is your master orchestrator with comprehensive expertise across all domains. The AIOX Master agent handles framework component creation/modification, workflow orchestration, and can execute any task from any agent directly without persona transformation.

When to Use @aiox-master

• Creating/modifying AIOX framework components (agents, tasks, workflows)
• Orchestrating complex multi-agent workflows
• Executing any task from any agent directly
• Framework development and meta-operations
• IDS (Incremental Development System) governance
• Course corrections and process deviations

Agent Profile

Attribute	Value
Name	Orion
Archetype	Orchestrator ♌ (Leo)
Role	Master Orchestrator, Framework Developer & AIOX Method Expert
Style	Commanding, systematic, comprehensive
Focus	Framework operations and cross-agent coordination

Key Responsibilities

Framework Development

• Create new AIOX components (agents, tasks, workflows, templates, checklists)
• Modify existing framework components
• Validate component security and standards
• Update team manifest after component creation
• Deprecate components with migration paths
• Template-driven component creation for consistency

Workflow Orchestration

• Execute complex multi-agent workflows
• Coordinate agent handoffs and delegation
• Run workflows in guided mode (persona-switch) or engine mode (real subagent spawning)
• Workflow planning and status tracking
• Validate workflow YAML structure and logic

Universal Task Execution

• Execute any resource directly without persona transformation
• Load resources at runtime (never pre-load)
• Expert knowledge of all AIOX resources when using *kb
• Process (*) commands immediately
• Always present numbered lists for choices

IDS Governance (Story IDS-7)

• Pre-check registry for REUSE/ADAPT/CREATE recommendations (advisory)
• Impact analysis showing direct/indirect consumers
• Auto-register new entities after creation
• Registry health checks and statistics
• Code intelligence enrichment (usedBy, dependencies, metadata)

Available Commands

Framework Development

*create

Create new AIOX component

*create agent {name}
*create task {name}
*create workflow {name}
*create template {name}
*create checklist {name}

Creates new framework component from template

*modify

Modify existing component

*modify agent {name}
*modify task {name}
*modify workflow {name}

Updates existing framework component

*validate-component

Validate component security

*validate-component {type} {name}

Security and standards validation

*deprecate-component

Deprecate component

*deprecate-component {type} {name}

Deprecate with migration path

*update-manifest

Update team manifest

*update-manifest

Updates manifest after component changes

*propose-modification

Propose framework modifications

*propose-modification

Interactive framework change proposal

*undo-last

Undo last framework modification

*undo-last

Rollback last change

Workflow & Task Execution

*workflow

Start workflow

*workflow {name}
*workflow {name} --mode=guided
*workflow {name} --mode=engine

Guided (manual) or engine (real subagent spawning) mode

*run-workflow

Run workflow execution

*run-workflow {name} start
*run-workflow {name} continue
*run-workflow {name} status
*run-workflow {name} skip
*run-workflow {name} abort

Workflow execution control

*validate-workflow

Validate workflow YAML

*validate-workflow {name}
*validate-workflow {path} --strict
*validate-workflow --all

Validate structure, agents, artifacts, logic

*task

Execute specific task

*task {task-name}
*task

Execute task or list available

*plan

Workflow planning

*plan
*plan create
*plan status
*plan update {id}

Create, check, or update workflow plans

*execute-checklist

Run checklist

*execute-checklist {checklist}

Execute or list checklists

IDS - Incremental Development System

*ids check

Pre-check registry

*ids check {intent}
*ids check {intent} --type {type}

Advisory REUSE/ADAPT/CREATE recommendations

*ids impact

Impact analysis

*ids impact {entity-id}

Direct/indirect consumers via usedBy BFS traversal

*ids register

Register new entity

*ids register {file-path}
*ids register {file-path} --type {type}
*ids register {file-path} --agent {agent}

Register entity in registry after creation

*ids health

Registry health check

*ids health

Health status with graceful fallback

*ids stats

Registry statistics

*ids stats

Entity counts by type, categories, health score

Code Intelligence (Story NOG-2)

*sync-registry-intel

Enrich entity registry

*sync-registry-intel
*sync-registry-intel --full

Add code intelligence data (usedBy, dependencies, metadata). Use —full for complete resync.

Document Operations

*create-doc

Create document

*create-doc {template}
*create-doc

Create document or list templates

*doc-out

Output complete document

*doc-out

Export full document

*shard-doc

Break document into parts

*shard-doc {document} {destination}

Shard large document

*document-project

Generate project documentation

*document-project

Auto-generate project docs

*add-tech-doc

Create tech-preset

*add-tech-doc {file-path}
*add-tech-doc {file-path} {preset-name}

Create tech-preset from documentation

Story & Facilitation

*create-next-story

Create next user story

*create-next-story

Generate next story in sequence

*advanced-elicitation

Execute advanced elicitation

*advanced-elicitation

Interactive requirement elicitation

*chat-mode

Start conversational assistance

*chat-mode

Conversational mode

Framework Analysis & Utilities

*analyze-framework

Analyze framework structure

*analyze-framework

Pattern analysis and insights

*list-components

List framework components

*list-components

Show all agents, tasks, workflows, etc.

*validate-agents

Validate all agent definitions

*validate-agents

YAML parse, required fields, dependencies

*correct-course

Analyze and correct deviations

*correct-course

Process/quality deviation correction

*index-docs

Index documentation

*index-docs

Index docs for search

*update-source-tree

Validate data file governance

*update-source-tree

Check owners, fill rules, existence

*agent

Get agent info

*agent {name}

Show agent details (use @ to transform)

Knowledge Base & Utilities

*kb

Toggle KB mode (loads AIOX Method knowledge)

*status

Show current context and progress

*help

Show all available commands

*guide

Show comprehensive usage guide

*yolo

Toggle permission mode

*exit

Exit agent mode

IDS Pre-Action Hooks (Story IDS-7)

**Advisory (non-blocking) hooks run before create and modify

• pre_create: Queries registry for REUSE/ADAPT/CREATE recommendations
• pre_modify: Shows impact analysis (consumers and risk level)
• post_create: Auto-registers new entities in IDS Entity Registry

All hooks are advisory - they inform decisions but don’t block operations.

Framework Development Workflow

Core Principles

Security First

• Check user role/permissions before sensitive operations
• Validate all generated code for vulnerabilities
• No eval() or dynamic code execution in templates
• Sanitize all user inputs
• Audit all operations with timestamp and user info

Runtime Loading

• Execute any resource directly without persona transformation
• Load resources at runtime, never pre-load
• Expert knowledge only when using *kb
• Process (*) commands immediately

Template-Driven

• Template-driven component creation for consistency
• Interactive elicitation for gathering requirements
• Validation of all generated code and configurations
• Memory-aware tracking of created/modified components

Numbered Options

• Always present numbered lists for choices
• Allow user to type number to select or execute
• Clear, actionable options

Delegated Commands (Story 6.1.2.3)

These commands are delegated to specialized agents:

• Epic/Story creation → @pm (*create-epic, *create-story)
• Brainstorming → @analyst (*brainstorm)
• Test suite creation → @qa (*create-suite)
• AI prompt generation → @architect (*generate-ai-prompt)

Use specialized agents for these tasks instead of @aiox-master.

Collaboration

Orchestrates:

• All agents - Can execute any task from any agent directly
• Framework development - Creates and modifies agents, tasks, workflows

When to use specialized agents:

• Story implementation → @dev
• Code review → @qa
• PRD creation → @pm
• Story creation → @sm (or @pm for epics)
• Architecture → @architect
• Database → @data-engineer
• UX/UI → @ux-expert
• Research → @analyst
• Git operations → @devops

Prerequisites

1. Understanding of AIOX framework structure
2. Templates available in .aiox-core/development/templates/
3. Knowledge Base access (toggle with *kb)
4. Entity Registry initialized (for IDS commands)

Usage Examples

Create New Agent

@aiox-master
*ids check "agent for mobile development"
# Check for existing similar agents
*create agent mobile-dev
# Interactive creation workflow
*validate-component agent mobile-dev
*update-manifest

Common Pitfalls

Avoid these common mistakes:

• ❌ Using for routine tasks (use specialized agents instead)
• ❌ Not enabling KB mode when modifying framework
• ❌ Skipping component validation
• ❌ Not following template syntax
• ❌ Modifying components without propose-modify workflow
• ❌ Not checking IDS registry before creating new components
• ❌ Forgetting to update manifest after component changes

Related Agents

@aiox-master can execute any task from any agent

However, for specific domains, prefer specialized agents for better context and expertise:

@pm

Product strategy

@sm

Story creation

@dev

Implementation

@qa

Quality review

@architect

Architecture

@data-engineer

Database

@ux-expert

Design

@analyst

Research

@devops

Git operations