Implement and Iterate Task
Implement tasks using appropriate coding agents with continuous build verification.
Usage: When user provides a task to implement
Instructions
- Respect AGENTS.md (and Claude.md if it exists)
- Use MCP servers to assist
- Delegate to appropriate coding agents when possible:
- Fast Coding Agent: Simple tasks, single-file edits
- Coding Agent: Medium complexity, multi-file changes
- Complex Coding Agent: High-complexity architectural work
- Spawn as many agents as needed, including using the fleet skill for parallel work
- Always use modern language syntax when possible
Iteration Loop
- Implement sub-task
- Build & Test
- Fix if needed (repeat 1-2)
- Code review Use multiple high-end models (Opus, Gemini Pro, Codex) to review code for quality, security, and best practices and alignment to the task. Iterate and fix (critical, high, medium) feedback.
- Next sub-task
Converted and distributed by TomeVault — claim your Tome and manage your conversions.