ultrathink-orchestrator

Auto-invokes ultrathink workflow for any work request (default orchestrator)

$ 安裝

git clone https://github.com/rysweet/MicrosoftHackathon2025-AgenticCoding /tmp/MicrosoftHackathon2025-AgenticCoding && cp -r /tmp/MicrosoftHackathon2025-AgenticCoding/docs/claude/skills/ultrathink-orchestrator ~/.claude/skills/MicrosoftHackathon2025-AgenticCoding

// tip: Run this command in your terminal to install the skill


name: ultrathink-orchestrator version: 1.0.0 description: Auto-invokes ultrathink workflow for any work request (default orchestrator) DEPRECATED: true deprecation_notice: | This skill is DEPRECATED as of 2025-11-26. Workflow selection is now handled directly in CLAUDE.md via the "MANDATORY: Workflow Selection" section. The indirection through skills has been eliminated to ensure consistent workflow enforcement. auto_activate: false priority: 5 triggers:

  • "implement"
  • "create"
  • "build"
  • "add"
  • "fix"
  • "update"
  • "refactor"
  • "design"

Ultrathink Orchestrator Skill

Purpose

This skill provides automatic orchestration for development and investigation tasks. It detects the task type from keywords and delegates to the appropriate workflow skill (investigation-workflow or default-workflow).

Auto-activation priority is LOW (5) to allow more specific skills to match first. When activated, this orchestrator selects between investigation-workflow and default-workflow based on the user's request keywords.

This skill acts as a thin wrapper around the canonical ultrathink command, following the amplihack pattern of single-source-of-truth for command logic.

Canonical Sources

This skill is a thin wrapper that references canonical sources:

  • Primary Command: .claude/commands/amplihack/ultrathink.md (278 lines)
  • Workflow Sources:
    • Development: .claude/workflow/DEFAULT_WORKFLOW.md
    • Investigation: .claude/workflow/INVESTIGATION_WORKFLOW.md

The canonical command contains complete task detection logic, complexity estimation, and orchestration patterns for both investigation and development workflows.

Execution Instructions

When this skill is activated, you MUST:

  1. Read the canonical command for task detection logic:

    Read(file_path=".claude/commands/amplihack/ultrathink.md")
    

    Note: Path is relative to project root. Claude Code resolves this automatically.

  2. Detect task type using keywords from the canonical command:

    • Investigation keywords: investigate, explain, understand, analyze, research, explore
    • Development keywords: implement, build, create, add feature, fix, refactor, deploy
    • Hybrid tasks: Both investigation and development keywords present
  3. Invoke appropriate workflow skill:

    • Investigation: Skill(skill="investigation-workflow")
    • Development: Skill(skill="default-workflow")
    • Hybrid: Both sequentially (investigation first, then development)
  4. Fallback if skill not found:

    • Read workflow markdown files directly
    • Investigation: .claude/workflow/INVESTIGATION_WORKFLOW.md
    • Development: .claude/workflow/DEFAULT_WORKFLOW.md
    • Follow workflow steps as specified

Why This Pattern

Benefits:

  • Single source of truth for orchestration logic in canonical command
  • No content duplication between command and skill
  • Task detection rules defined once, maintained once
  • Changes to ultrathink command automatically inherited by skill

Trade-offs:

  • Requires Read tool call to fetch canonical logic
  • Slight indirection vs. inline implementation

This pattern aligns with amplihack philosophy: ruthless simplicity through elimination of duplication.

Related Files

  • Canonical Command: .claude/commands/amplihack/ultrathink.md
  • Development Workflow Skill: .claude/skills/default-workflow/
  • Investigation Workflow Skill: .claude/skills/investigation-workflow/
  • Canonical Workflows:
    • .claude/workflow/DEFAULT_WORKFLOW.md
    • .claude/workflow/INVESTIGATION_WORKFLOW.md