Find agent skills by outcome

3. Matte black stealth armor

sitting upright holding a folded newspaper open in both hands

multimodal Reasoning: If asked

explain the why behind a step using the 2025 evidence (e.g.

- A05 Security Misconfiguration: DEBUG=True in prod

verbose errors

application security (OWASP

STRIDE threat modeling) and code architecture

stack: ${stack:React 19

Next.js App Router

in React

Tailwind CSS

- Verify tenant isolation with concrete test scenarios (e.g.

user A requests /api/metrics/?tenant_id=B)

in codebase

.env files are gitignored

and verify DEBUG=False

SECURE_SSL_REDIRECT=True

Deep Learning Loop

# Deep Learning Loop System v1.0

- ⚪ Low: Only in code

not user-facing (test fixtures

For the top 5 fixes

provide copy-paste-ready code or configuration.,FALSE,TEXT,gokbeyinac

For each metric

explain:

If no CLAUDE.md exists yet: Tell the prompt to generate a minimal token section at the top of the handoff document covering only the tokens needed for this specific handoff. It won't be a full design system

but it prevents hardcoded values.

After implementation

the AI agent (or human developer) should verify:

- icon: ImageOff from lucide-react

text-gray-400

Your output will be read primarily by AI coding agents (Claude Code

Cursor)

- Primary: AI coding agents (Claude Code

Cursor

### A structured handoff document optimized for AI implementation agents (Claude Code

Cursor

await page.screenshot({ path: 'debug.png'

fullPage: true });

1. Inspect the existing codebase

components

Dependencies: React 18

Next.js 14

static regeneration

utilizing Next.js App Router with React Server

Explain why decisions were made

what alternatives were considered