orchestration-commands

name	orchestration-commands
description	Registry and index of all orchestration command skills
category	Orchestration Framework
metadata	{"version":"1.0","source":"Agent33","original_command":"COMMAND_REGISTRY","tags":["orchestration","commands","registry","index"]}

Orchestration Command Registry

This registry defines all available orchestration command skills. Each command is a standalone skill that can be invoked independently.

Available Commands

Command	Purpose	Skill Path
orch-status	Review runtime state and surface blockers	orch-status/SKILL.md
orch-tasks	List and manage open tasks with priorities	orch-tasks/SKILL.md
orch-verify	Capture verification evidence for current task	orch-verify/SKILL.md
orch-handoff	Generate session wrap summaries	orch-handoff/SKILL.md
orch-plan	Create approval-gated implementation plans	orch-plan/SKILL.md
orch-review	Trigger code review with risk routing	orch-review/SKILL.md
orch-tdd	RED/GREEN/REFACTOR workflow entry point	orch-tdd/SKILL.md
orch-build-fix	Diagnose and fix build/test failures	orch-build-fix/SKILL.md
orch-refactor	Dead code cleanup and refactoring	orch-refactor/SKILL.md
orch-docs	Documentation synchronization	orch-docs/SKILL.md
orch-e2e	Generate and run E2E test suites	orch-e2e/SKILL.md

Schema

Each command skill follows this schema:

id: string           # Unique identifier (orch-prefixed)
name: string         # Display name
description: string  # Brief purpose statement
triggers:            # What invokes this command
  - manual           # Operator-initiated
  - scheduled        # Time-based
  - event            # Triggered by system event
inputs:              # Required context
  - document: string # Input path or description
    required: bool   # Whether mandatory
outputs:             # Produced artifacts
  - document: string # Output path or pattern
    action: string   # create | update | append

Command Conventions

Invocation

Commands are invoked by referencing the skill name:

orch-{command} [required-args] [optional-args]

Standard Outputs

All commands should produce:

1. Primary artifacts (code, docs, tests)
2. Evidence of execution
3. Task tracking update (if applicable)

Error Handling

• Commands should fail gracefully with clear messages
• Partial progress should be captured
• Escalation path should be clear

Adding New Commands

1. Create a new subdirectory under commands/ named orch-<id>/
2. Add a SKILL.md following the frontmatter schema
3. Add an entry to this registry table
4. Document any new artifacts or dependencies

Related Skills

• ../handoff-protocol/ - Handoff protocol definitions (when available)
• ../../GENERAL CODING WORKFLOWS/session-wrap/ - Session wrap workflow