From lighterpowers
Use when executing an implementation plan. Offers batch, validated-batch, subagent, or one-shot approaches.
How this skill is triggered — by the user, by Claude, or both
Slash command
/lighterpowers:execute-planThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Execute an implementation plan. First, determine which execution style to use.
Execute an implementation plan. First, determine which execution style to use.
COMPULSORY: Always ask the user which approach they want. Never skip this question.
docs/lighterpowers/plans/feature.md)--batch-size=N: Optional, passed to batch-development (default: 3)Use AskUserQuestion with these options:
Question: "How would you like to execute this plan?"
| Option | Label | Description |
|---|---|---|
| A | Batch (human checkpoints) (Recommended) | Execute tasks in batches, pause for your feedback after each batch. You stay in control. |
| B | Validated Batch (parallel validation) | Intelligent batching based on buildability + 3 parallel validators after each batch. Best balance of speed and quality. |
| C | Subagent (automated reviews) | Fresh subagent per task with automated spec + code quality reviews. Faster, less interaction. |
| D | One Shot (not recommended) | Execute ALL tasks without pauses, then build+test. Requires solid plan. No checkpoints until build completes. |
This question is COMPULSORY. Never skip it, never assume a default.
Skill(lighterpowers:batch-development, args: "<plan-path> --batch-size=N")Skill(lighterpowers:validated-batch-development, args: "<plan-path>")Skill(lighterpowers:subagent-driven-development, args: "<plan-path>")Skill(lighterpowers:one-shot-development, args: "<plan-path>")Pass the plan path as the argument to the skill.
| Violation | Why It's Critical | Recovery |
|---|---|---|
| Skipping the choice question | User loses control over execution style | Present AskUserQuestion |
| Assuming a default | Different users want different approaches | Always ask |
| Not passing batch-size arg | User's preference ignored | Parse and forward --batch-size |
Guides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.
npx claudepluginhub bradwindy/lighterpowers