The result? Decent code, but not quite right. Another attempt with more details. Still not there.

Then I discovered Claude agents. One line: "Use flutter-developer agent." Done. Production-ready code.

The Problem Nobody Talks About

You know that feeling when you ask Claude to write Flutter code and it gives you something that works but isn't right? No proper state management. No error handling. Just a basic StatefulWidget that makes you cringe.

Why This Happens

I asked Claude directly. The response was eye-opening:

Specialized agents come with:

• Pre-configured prompts for specific frameworks (Flutter even provides official AI rules)
• Built-in best practices and conventions
• Industry standards by default

Generic Claude is intentionally flexible:

• No framework-specific defaults
• Treats each request neutrally
• Defaults to basic, functional code

The Bigger Picture

This isn't a limitation—it's actually a feature. Claude's flexibility lets you specify exactly what standards and patterns you want, rather than being locked into one opinionated setup. The trade-off is that you need to be more explicit about your requirements.

But who has time to be explicit about everything, every time?

Real-World Example

Generic Claude:

Write a Flutter login screen with:
- BLoC pattern
- Clean architecture
- Error handling
- Form validation
- Unit tests

Result: Basic implementation. Missing details. Needs work.

With Agent:

@flutter-developer create a login screen

Result: Everything above, plus proper folder structure, dependency injection, loading states. Production-ready.

The Agents I Actually Use

Daily:

• @flutter-developer - Flutter features
• @debugger - Bug fixes
• @test-automator - Test coverage
• @code-reviewer - Quality checks

Specialized:

• @security-auditor - Security review
• @performance-engineer - Optimization
• @database-optimizer - Query tuning

The Quality Gap

Generic Claude: Basic code that works but needs refactoring

Agents: Production-ready with proper architecture and tests

When to Use What

Generic Claude: Quick explanations, one-liners, small snippets

Agents: Real features, production code, complex implementations

Common Mistakes

1. Using wrong agent (@javascript-pro for React instead of @frontend-developer)
2. Over-controlling the agent instead of letting it use its expertise
3. Not using review agents before commits

The Bottom Line

AI models work best with clear context about structure, standards, and requirements. Agents have all that built in.

Why write detailed prompts when you can call the specialist who already knows?

Generic Claude is the brilliant generalist. Agents are the specialists. Both have their place. But when you need production code, you know which to choose.

Next time you're writing a long prompt, ask: Is there an agent for this?

There probably is. And it knows the best practices better than your prompt ever will.

Here's everything useful I learned, written to stay relevant regardless of when you're reading this.

1️⃣ Do You Even Need It?

You NEED it if:

• You have spouse/kids
• Parents depend on your income
• You have loans (home/personal/education)
• You're the sole breadwinner
• Family can't maintain lifestyle without your income

You DON'T need it if:

• Single with no dependents
• Spouse earns enough
• Parents are financially independent
• Your wealth > 25x annual expenses
• Zero debts or liabilities

Bottom line: Even one dependent = Get term insurance this week

2️⃣ How Much Coverage?

The Formula

Coverage = Annual Expenses × 10-15 + All Debts + Future Goals - Liquid Assets

Real Example (2024 Numbers)

Let's say you're 32, married with 1 kid, earning ₹12 lakhs/year:

*Using 12x multiplier for moderate coverage. Can go up to 15x for safer coverage.

⚠️ These numbers will change! In 2030, the same ₹50,000/month lifestyle might cost ₹75,000/month. That's why we use multipliers (10-15x) not fixed amounts. The formula stays same, numbers change with inflation.

Quick Rules of Thumb

• Minimum: 10x annual income
• Comfortable: 15x annual income
• Conservative: 20x annual income
• Review: Every 3 years or major life event

3️⃣ Policy Duration

Default: Till retirement age (typically 60-65)

Stop paying when:

• Your wealth > coverage amount
• No more dependents
• Achieved 25x annual expenses saved (FIRE number)

⚠️ Premium increases with age: 25 yo vs 35 yo = 2x difference, 25 yo vs 45 yo = 4-5x difference

4️⃣ Which Plan to Choose?

Get This

✅ Basic Term Insurance - Pure death benefit, nothing else

Avoid These

❌ Money-back plans ❌ Investment + Insurance combos ❌ Whole life insurance ❌ Any plan promising "returns"

How to Save Money

Payment frequency matters:

• Yearly: Best rate (baseline)
• Half-yearly: 2% extra
• Quarterly: 4% extra
• Monthly: 5% extra

Buy online: Save 5-10% (no agent commission)

5️⃣ Choosing the Insurer

Must Check These Numbers

Red Flags 🚩

• Agent pushing investment products
• "Guaranteed returns" promises
• Pressure tactics ("offer ends today")

6️⃣ Skip the Riders

My approach: Basic term + Comprehensive health insurance

7️⃣ Avoid Claim Rejection

Never Hide (Even Small Things)

• Smoking (even occasionally)
• Drinking (even socially)
• Medical history (even minor)
• Family health issues
• Actual income
• Previous policies

Top Rejection Reasons

1. Lied about lifestyle - 35% of rejections
2. Hidden medical history - 25%
3. Policy lapsed - 15%
4. Wrong income declared - 10%
5. Death in first 2 years - Contestability period

8️⃣ After You Buy

Do immediately:

• Add nominee with full details
• Tell nominee about the policy
• Save policy number in multiple places
• Set auto-pay for premiums
• Upload documents to cloud

Keep ready for claims:

• Death certificate
• Policy document
• Nominee KYC
• Medical records
• FIR (if accident)

9️⃣ My Decision Framework (2020)

The Trade-off I Made

• Max Life quoted ₹11,000/year (20% cheaper)
• HDFC Life was ₹14,000/year
• Chose HDFC for better claim settlement (98% vs 94%)
• Extra ₹3,000/year = ₹90,000 over 30 years
• Worth it? For peace of mind, absolutely

📝 Note: These premiums are from 2020 for a 30-year-old non-smoker. Your premium will differ based on age, health, and year of purchase.

How to Decide in Any Year

1. List top 5 insurers by claim settlement ratio
2. Get quotes from all online
3. Pick the one with best claims record (not cheapest)
4. Buy the largest coverage you can afford
5. Pay annually if possible

The 30-Second Summary

1. Need it? Yes, if anyone depends on you
2. How much? Annual expenses × 10-15 + debts
3. Which type? Basic term plan, no riders
4. Where? Online, direct from insurer
5. When? Now. Every year = higher premium

Final thought: Term insurance is boring. That's the point. Get it, forget it, build wealth.

My money. Stuck. In a "safe" fund.

The Setup

Like many developers, I'm good with logic but cautious with money. Stock markets? Too volatile. Crypto? Don't even start. I wanted something boring and predictable - better than my 3.5% savings account but without the drama.

Enter Franklin Ultra Short Duration Fund. The pitch was perfect:

• Better returns than FDs (7-8% vs 5-6%)
• "Ultra short" = super safe, right?
• Liquid whenever you need it
• Franklin Templeton - big global brand, 25+ years in India

I did my homework. Or so I thought. Star ratings, past returns, fund size - all looked solid. Moved a decent chunk of my emergency fund there in 2019.

The Crash

Then COVID hit. And Franklin did something nobody expected.

They didn't just lose value. They literally shut down. "Winding up" they called it. Six debt schemes. ₹26,000 crores locked. My fund included.

My money wasn't just stuck. It was frozen with no timeline for return.

The Lessons Nobody Tells You

1. "Low Risk" Doesn't Mean "No Risk"

My fund was investing in corporate bonds. When companies started struggling during COVID, those bonds became worthless paper. The fund couldn't sell them even if they wanted to.

It's like having a perfectly good laptop that nobody wants to buy. The value exists only if someone's willing to pay.

2. Liquidity Can Vanish Overnight

"You can withdraw anytime" - except when you can't. When everyone rushes for the exit, the door gets jammed. Too many withdrawal requests, and the system simply stops.

3. Big Names Can Fall Too

Franklin Templeton. Since 1996 in India. ₹1 lakh crore under management.

If they can freeze funds, anyone can. Brand value means nothing when fundamentals break.

What Actually Happened to My Money

The waiting game began. Monthly updates. Legal proceedings. Partial payments.

• First payment: 15% after 8 months
• Second payment: 20% after a year
• Trickles continued for 2+ years

Final recovery: About 92% of my investment. No returns. Just principal, minus 8%.

Meanwhile, my friend's boring FD gave him 6% annually with zero drama.

The Real Irony

As a developer, I live by the rule: never trust a single point of failure. Yet with money, I trusted a single fund manager's decisions.

I spent more time reviewing a ₹500 gadget on Amazon than researching where to park lakhs.

What I Do Now

I keep things simple. Emergency money stays in boring places - savings accounts and FDs. No fancy schemes, no chasing extra returns.

For growth? I finally made peace with equity markets. Index funds mostly. The volatility I was avoiding? Turns out it's more predictable than "safe" debt funds freezing overnight.

The irony isn't lost on me.

The Real Lesson

Debt funds aren't evil. But they're not fixed deposits with better returns. They're market-linked products with real risks.

That marketing term "Ultra Short Duration"? It refers to bond maturity, not your access to money.

Would I invest in debt funds again? Yes, but:

• Never emergency money
• Never more than 10% in one fund
• Only money I can forget for 3+ years
• Always checking what they're actually buying

Looking Back

2020 taught everyone different lessons. Mine was simple: There's no free lunch in finance.

Every percent above FD rates comes with proportional risk. The question isn't whether risk exists - it's whether you understand and accept it.

Today, my frozen funds story is a dinner conversation starter. Back then, it was sleepless nights wondering if I'd ever see that money again.

The 8% I lost? Consider it tuition fees for Financial Reality 101.

After years of shipping code, I'm finally shipping words.

Who Am I

Frontend Tech Lead based in Singapore, building mobile applications for global markets. I write code by day and analyze markets by night.

What I Write About

• Tech: Development insights, architecture patterns, and engineering lessons learned
• Finance: Market observations from a developer's perspective
• Product: Building things that matter

Why This Blog

Code tells computers what to do. Words tell humans why it matters.

This is where I share both.