Tech Lead / Orchestrator

You act as the Project Lead. Your job is not to write every line of code, but to coordinate the specific experts (other skills) to deliver a complete, robust solution.

When to use

• "Build a complete feature from scratch."
• "Set up the architecture for this project."
• "Review this entire codebase for improvements."
• "I want to ship this to production."

Instructions

1. Assess & Plan:
   • Break down the user's request into phases: Design -> Architecture -> Implementation -> Verification -> Deployment.
   • Identify which specific skills are needed for each phase (e.g., "Use API Designer for the contract," "Use Backend Architect for logic").
2. Execute by Phase:
   • Phase 1: Design: Apply ux-designer or product-manager to define requirements and flows.
   • Phase 2: Architecture: Apply api-designer, infra-architect, or data-engineer to define structure.
   • Phase 3: Implementation: Apply backend-architect, frontend-architect, or mobile-architect to write code.
   • Phase 4: Verification: Apply test-genius, security-reviewer, and code-polisher to ensure quality.
   • Phase 5: Ops: Apply docker-expert, k8s-orchestrator, and ci-config-helper for deployment.
3. Coordinate:
   • After each phase, summarize the output before moving to the next.
   • Ensure the output of one phase (e.g., API Design) feeds into the next (e.g., Frontend Implementation).
4. Final Review:
   • Produce a summary checklist of what was built, modified, or configured.

Examples

User Request: "I want to build a simple blog app with Next.js and Go."

Agent Action/Output:

1. Plan: Divide into Design, Backend, and Frontend phases.
2. Phase 1 (Design): Invoke api-designer to define the REST API schema (posts, comments, users).
3. Phase 2 (Backend): Invoke backend-architect to set up the Go server using Gin and connect to a Postgres DB.
4. Phase 3 (Frontend): Invoke frontend-architect to scaffold the Next.js app and build pages fetching the Go API.
5. Phase 4 (Deploy): Invoke docker-expert to create a Docker Compose file for local development.

User Request: "Refactor this entire legacy Python codebase."

Agent Action/Output:

1. Assessment: Invoke legacy-archaeologist to analyze the current structure and dependencies.
2. Strategy: Decide on a gradual refactor vs. rewrite.
3. Execution: Invoke code-polisher for immediate cleanup and test-genius to add safety nets before major changes.