What problem does it solve? Turning a written implementation plan into actual code requires tracking which tasks are done, resuming after interruptions, and keeping plan checkboxes in sync with reality. This Skill automates that execution loop so multi-session coding work never loses progress. ## Core Features & Use Cases - Sequential Task Execution: Reads a plan file, works through tasks one at a time, marks them in_progress/completed, and updates - [ ] checkboxes to - [x] immediately after each task. - Session Persistence & Recovery: Rebuilds context from git state and plan files after /clear or a new session, reconciling TaskList statuses with plan checkboxes before continuing. - Plan Discovery & Routing: Auto-detects branch-named plans, fast plans, and fix plans, supports @plan-file overrides, --list mode, and redirects fix plans to the fix workflow. - Handoff & Artifact Sync: Syncs status with a Handoff coordinator via MCP tools, updates DESCRIPTION/ARCHITECTURE/ROADMAP artifacts when the stack or structure changes, and handles commit checkpoints and worktree merges. - Use Case: After running /aif-plan full Add user authentication, invoke this Skill across multiple sessions to implement all six tasks, commit at checkpoints, and finish with verification and roadmap milestone updates. ## Quick Start Ask the AI to run /aif-implement to continue executing the current plan from the next incomplete task.