request-handler

Request Handler Workflow

Safety Notice

This listing is imported from skills.sh public index metadata. Review upstream SKILL.md and repository scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "request-handler" with this command: npx skills add duck4nh/antigravity-kit/duck4nh-antigravity-kit-request-handler

Request Handler Workflow

When receiving a user request, follow this process:

Step 1: Classify the Task

Identify which of the 4 categories the request belongs to:

Icon Type Keywords to Detect

CONSULT "should", "recommend", "compare", "suggest", "advice"

BUILD "create", "make", "build", "add", "implement", "write"

DEBUG "error", "bug", "not working", "wrong", "fix"

OPTIMIZE "slow", "refactor", "clean", "improve", "optimize"

Note: If unclear → Ask the user before proceeding.

Step 2: Execute Based on Mode

CONSULT Mode

  • Clarify context & constraints

  • Provide 2-3 options with clear trade-offs

  • Recommend the optimal option with reasoning

  • WAIT for confirmation before coding

BUILD Mode

  • Confirm scope & acceptance criteria

  • Propose file/component structure

  • Code in order: Types → Logic/Hooks → UI → Styles

  • Run checklist before delivery

DEBUG Mode

  • Gather info: what, where, when

  • Analyze root cause

  • Propose fix + explanation

  • Suggest prevention measures

OPTIMIZE Mode

  • Measure baseline

  • Identify main bottlenecks

  • Propose improvements + predict results

  • Refactor + compare before/after

Step 3: Pre-Delivery Checklist

Code Quality:

  • No any types

  • No hardcoded magic numbers/strings

  • Proper error handling

  • Clear variable/function naming

Structure:

  • Correct folder structure

  • Consistent naming convention

  • Split files appropriately (< 200 lines/file)

UI/UX (if applicable):

  • Follows Design System

  • Responsive, mobile-first

  • Loading/Error/Empty states

Tips

  • Don't expand scope unilaterally

  • Don't use any types

  • Ask when requirements are unclear

  • Comment complex logic

  • Prioritize: Readability → Performance → Cleverness

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

Automation

bash-automation

No summary provided by upstream source.

Repository SourceNeeds Review
General

linux-server-expert

No summary provided by upstream source.

Repository SourceNeeds Review
General

pentest-expert

No summary provided by upstream source.

Repository SourceNeeds Review