Execute

SkillFiles & storage

Execute a plan file — parse phases, spawn parallel subagents, handle failures.

Available today. Use it from your connected AI after setup.

No other account needed.

Connect ahel once, and every AI you use reads what you have installed.

Then ask your AI: use the Execute skill

What this skill tells your AI

The instructions your AI receives, as published by simonepri/ifttt-lint in .agents/skills/execute/SKILL.md and read by ahel’s review.

Execute a plan file produced by /plan. $ARGUMENTS is the path to the plan file.

Parse: read the plan file. Extract the phases and tasks. Each phase contains independent tasks that can run in parallel.
Execute phase by phase: for each phase:
- Independent tasks (touching separate files/modules): spawn a subagent per task. Give each subagent the plan file path, its assigned task, and the interfaces it needs to implement.
- Interdependent tasks (shared state, cross-cutting concerns): use an agent team instead — agents share a task list and coordinate directly, which avoids integration conflicts.
- Wait for all agents in the phase to complete before starting the next phase.
Handle failures: if a subagent gets stuck or goes off track, kill it, revert only that task's changes (not the entire branch — other tasks in the phase may have succeeded), and spawn a fresh one with adjusted instructions. If a task fails repeatedly (3 attempts), stop and report the failure — don't block other phases that don't depend on it.
Report: return a summary of completed tasks and any failures.

Advanced