From magic-powers
Use when executing implementation plans with independent tasks in the current session
npx claudepluginhub kienbui1995/magic-powers --plugin magic-powersThis skill uses the workspace's default tool permissions.
Execute plan by dispatching fresh subagent per task, with review after each.
Generates design tokens/docs from CSS/Tailwind/styled-components codebases, audits visual consistency across 10 dimensions, detects AI slop in UI.
Records polished WebM UI demo videos of web apps using Playwright with cursor overlay, natural pacing, and three-phase scripting. Activates for demo, walkthrough, screen recording, or tutorial requests.
Delivers idiomatic Kotlin patterns for null safety, immutability, sealed classes, coroutines, Flows, extensions, DSL builders, and Gradle DSL. Use when writing, reviewing, refactoring, or designing Kotlin code.
Execute plan by dispatching fresh subagent per task, with review after each.
Why subagents: Fresh context per task = focused work. Precisely crafted instructions = reliable results. Your context preserved for coordination.
Core principle: Fresh subagent per task + review = high quality, fast iteration.
vs. executing-plans: Same session, fresh context per task, review between tasks, faster iteration.
Dispatch implementer subagent with:
Implementer works: implements, tests, commits
Dispatch reviewer agent (Haiku — fast, cheap):
Handle review results:
Mark task complete in TodoWrite
| Role | Agent | Model | Why |
|---|---|---|---|
| Implementation | (default subagent) | Sonnet | Good balance of quality/cost |
| Code review | reviewer | Haiku | Fast, cheap, read-only |
| Complex design questions | architect | Opus | Deep reasoning needed |
When dispatching an implementer subagent, provide:
## Task
[Exact task from plan]
## Context
- Working directory: [path]
- Relevant files: [list]
- Dependencies: [what previous tasks created]
## Requirements
- Follow TDD: write test first, verify it fails, implement, verify it passes
- Commit after each step
- Run full test suite before reporting done
## Do NOT
- Modify files outside the task scope
- Skip tests
- Change architecture decisions from the plan
After all tasks: