Generate architecture guidelines for the project. Analyzes tech stack from DESCRIPTION.md, recommends an architecture pattern, and creates .ai-factory/ARCHITECTURE.md. Use when setting up project architecture, asking "which architecture", or after /aif setup.
Fix a specific bug or problem in the codebase. Supports two modes - immediate fix or plan-first. Without arguments executes existing FIX_PLAN.md. Always suggests test coverage and adds logging. Use when user says "fix bug", "debug this", "something is broken", or pastes an error message.
Execute implementation tasks from the current plan. Works through tasks sequentially, marks completion, and preserves progress for continuation across sessions. Use when user says "implement", "start coding", "execute plan", or "continue implementation".
Refine an existing implementation plan with a second iteration. Re-analyzes the codebase for gaps, missing tasks, and wrong dependencies. Use after /aif-plan or to improve an /aif-fix plan. Optional +check flag validates refinements via a fresh-context subagent.
Run a strict multi-iteration Reflex Loop with phases (PLAN, PRODUCE||PREPARE, EVALUATE, CRITIQUE, REFINE) to improve an artifact until quality gates pass or iteration limits are reached. Use when user asks for iterative refinement, quality-gated generation, or "generate -> critique -> refine" loops.
Plan implementation for a feature or task. Two modes — fast (single quick plan) or full (richer plan with optional git branch/worktree flow). Use when user says "plan", "new feature", "start feature", "create tasks".
QA workflow for testing a feature or task implementation. Analyzes changes, produces test plans, and describes concrete test scenarios. Use when user says "test this", "write test plan", "what should I test", or "QA this branch".
Perform code review on staged changes or a pull request. Checks for bugs, security issues, performance problems, and best practices. Use when user says "review code", "check my code", "review PR", or "is this code okay". Optional +check flag validates findings via a fresh-context subagent.