forge-skill-orchestrator

To auto-discover domain skills for AI workflow tasks, use an orchestrator that detects relevant skills from project and global directories and loads them automatically at task start, reducing manual configuration errors.

Phase-aware skill orchestration in Claude Code coordinates multiple sub-skills across design, spec, implement, review, test, and document phases, automatically selecting the correct methodology and context files for each specific workflow stage.

To coordinate multiple domain skills across workflow phases, proactively invoke a phase-aware orchestrator at task start to ensure standardized prompts for sub-agents and seamless integration from brainstorming through documentation.

Yes, phase detection can select methodologies for brainstorming, implementation, and testing by automatically determining the correct workflow stage and loading the appropriate domain skills and context files for each phase.

When domain skills lack phase-aware context files, the orchestrator falls back to loading the complete SKILL.md contents, ensuring full guidance and standardized prompts for sub-agents are still available during the workflow.

No, manual configuration of sub-agent prompts is unnecessary for end-to-end AI task orchestration because the orchestrator automatically generates standardized prompts and integrates domain auto-discovery to coordinate sub-skills across all phases.