Skip to main content
OrchestKit v6.7.1 — 67 skills, 38 agents, 77 hooks with Opus 4.6 support
OrchestKit

24 Commands You Can Invoke

Every /ork:X command skill in OrchestKit -- grouped by category with descriptions and usage examples.

What Are Command Skills?

Command skills are the 24 skills with user-invocable: true in their frontmatter. You trigger them by typing /ork:skill-name in Claude Code. Each one is a structured workflow that guides Claude through a multi-step process, often spawning parallel agents and pulling in reference skills automatically.

Command skills differ from reference skills in three ways:

  1. You invoke them explicitly with the /ork: prefix.
  2. They define workflows, not just knowledge -- step-by-step processes with decision points.
  3. They compose other skills via the skills: frontmatter field, pulling in reference skills as needed.

Implementation and Development

These commands handle the core development cycle: building features, fixing bugs, and managing code quality.

CommandDescriptionComplexityExample
/ork:implementFull-power feature implementation with parallel subagents. Asks for scope (full-stack, backend-only, frontend-only, quick prototype) and constraints before starting.medium/ork:implement user authentication
/ork:fix-issueSystematic GitHub issue resolution with hypothesis-based root cause analysis, similar issue detection, and prevention recommendations.medium/ork:fix-issue 123
/ork:verifyComprehensive verification with parallel test agents. Validates implementations across unit tests, integration tests, security, and type safety.low/ork:verify
/ork:exploreDeep codebase exploration with parallel agents. Maps architecture, discovers patterns, generates ASCII diagrams, and produces health assessments.low/ork:explore
/ork:brainstormingDesign exploration with parallel agents. Generates ideas, compares alternatives, and produces structured recommendations.medium/ork:brainstorming real-time notifications
/ork:assessRates quality 0-10 with structured pros/cons analysis. Works on code, designs, approaches, or any artifact.low/ork:assess this API design
/ork:assess-complexityAssesses task complexity with codebase metrics. Determines if a task needs breakdown into subtasks.medium/ork:assess-complexity add payment processing

Git and GitHub

Commands for version control and pull request workflows.

CommandDescriptionComplexityExample
/ork:commitCreates commits with conventional format and validation. Analyzes staged changes, generates a message, and runs pre-commit checks.low/ork:commit
/ork:create-prCreates GitHub pull requests with validation. Runs security scanning, code review, and type checking before opening the PR.low/ork:create-pr
/ork:review-prPR review with parallel specialized agents. Checks security, test coverage, type safety, and code quality simultaneously.medium/ork:review-pr 456
/ork:git-recoveryRecovery from git mistakes. Guides you through undoing commits, recovering branches, or fixing history using reflog.medium/ork:git-recovery

Memory and Knowledge

Commands for storing and retrieving decisions, patterns, and context across sessions.

CommandDescriptionComplexityExample
/ork:rememberStores decisions and patterns in the knowledge graph. Supports --success, --failed, and --agent flags for categorization.low/ork:remember --success cursor pagination scales well
/ork:memoryRead-side memory operations: search, load, sync, history, and visualization. Subcommands: search, load, sync, history, viz, status.low/ork:memory search "pagination"

Plugin Management

Commands for managing OrchestKit itself -- health checks, configuration, skill evolution, and feedback.

CommandDescriptionComplexityExample
/ork:doctorOrchestKit health diagnostics. Validates permissions, hooks, skills, agents, and memory system status.low/ork:doctor
/ork:configureConfigures OrchestKit settings. Interactive wizard for MCP servers, plugin options, and preferences.low/ork:configure
/ork:helpOrchestKit skill directory. Lists available skills grouped by category with descriptions.low/ork:help
/ork:feedbackManages OrchestKit feedback system. Submit feedback or view usage analytics with privacy controls.low/ork:feedback
/ork:skill-evolutionEvolves skills based on usage patterns. Analyzes skill effectiveness and suggests improvements or rollbacks.medium/ork:skill-evolution

Specialized Workflows

Commands for specific domains -- data curation, parallel development, competitive intelligence, platform upgrades, and video production.

CommandDescriptionComplexityExample
/ork:add-goldenAdds documents to golden dataset with validation. Includes quality scoring and bias detection for test data curation.medium/ork:add-golden
/ork:worktree-coordinationCoordinates multiple Claude instances across git worktrees. Manages locking for parallel development.low/ork:worktree-coordination
/ork:competitive-monitoringTracks competitor page changes over time. Captures snapshots, detects diffs, and alerts on significant changes. Supports Tavily for URL discovery.medium/ork:competitive-monitoring
/ork:upgrade-assessmentAssesses platform upgrade readiness for Claude model and CC version changes. Full compatibility and migration analysis.max/ork:upgrade-assessment
/ork:demo-producerCreates polished demo videos for skills, tutorials, and CLI demonstrations. Orchestrates recording, narration, and composition.low/ork:demo-producer

How Command Skills Compose Reference Skills

Every command skill lists reference skills in its frontmatter skills: field. When you invoke the command, those reference skills are automatically loaded into context. For example, /ork:implement pulls in:

skills:
  - api-design-framework
  - react-server-components-framework
  - type-safety-validation
  - unit-testing
  - integration-testing
  - explore
  - verify
  - memory
  - worktree-coordination

This means you do not need to remember which reference skills exist. The command skill knows what knowledge it needs and loads it automatically.


Invocation Patterns

Basic invocation

/ork:implement user authentication with OAuth

With GitHub issue reference

/ork:fix-issue 123

With flags (memory commands)

/ork:remember --success --category database pgvector handles 10M vectors well
/ork:memory search "caching strategy"

Chaining commands in a session

/ork:explore                          # Understand the codebase
/ork:implement search feature         # Build the feature
/ork:verify                           # Validate the implementation
/ork:commit                           # Commit the changes
/ork:create-pr                        # Open a pull request

What's Next

Edit on GitHub

Last updated on