In partnership with
MORNING COFFEE
How to Give Claude Superpowers (Without Touching a Config File)
Anthropic dropped Desktop Extensions on Monday. The feature was quiet—no press release, no launch party—but it's about to change how non-technical users interact with AI forever.
The numbers: 5 minutes to full setup (down from 45+ minutes of terminal commands). One-click installation (versus manual JSON editing). Zero Node.js knowledge required (previously a hard prerequisite).
Here's what happened. The Model Context Protocol (MCP) launched late 2024 as the "USB standard for AI"—a way to connect Claude to your files, databases, APIs, and workflows. The problem? Installation required editing configuration files buried in system folders, running terminal commands, and troubleshooting cryptic error logs. Non-developers bounced hard.
Desktop Extensions solve this entirely. Now you can extend Claude's capabilities the same way you install a Chrome extension: browse a directory, click "Install," and you're done.
Three signals builders should parse:
The Extension Directory is curated—Anthropic reviews submissions before publishing. This signals quality control matters more than marketplace volume. Getting listed early creates distribution moat.
Enterprise controls are baked in—Team and Enterprise admins can create allowlists, upload custom extensions, and manage access. This isn't just consumer-facing—it's an enterprise sales accelerator.
The spec is open-source—Anthropic published the MCPB format, schemas, and toolchain publicly. They want other AI desktop apps to adopt it. This is a standards play, not a walled garden.
Here's how to set it up (no code required):
Step 1: Download Claude Desktop from claude.ai/download (macOS or Windows).
Step 2: Open Claude Desktop → Click the hamburger menu (☰) → Settings → Extensions.
Step 3: Click "Browse extensions" and select any Anthropic-reviewed tool (Filesystem, GitHub, Slack, etc.).
Step 4: Click "Install Extension" → Follow the prompts → Grant permissions when asked.
Step 5: Restart Claude Desktop. Look for the hammer icon (🔨) in the bottom-right of your input box—this confirms your tools are active.
That's it. Five steps. No terminal. No JSON files. No Stack Overflow rabbit holes.
The market reaction hasn't fully priced this in. MCP servers were powerful but inaccessible—Desktop Extensions remove the friction entirely. When your non-technical marketing manager can give Claude access to Google Drive in 60 seconds, adoption curves change.
The question: If AI assistants can now read your files, search your email, and execute workflows with one click, what's the last job that requires human copy-paste?
Hiring in 8 countries shouldn't require 8 different processes
This guide from Deel breaks down how to build one global hiring system. You’ll learn about assessment frameworks that scale, how to do headcount planning across regions, and even intake processes that work everywhere. As HR pros know, hiring in one country is hard enough. So let this free global hiring guide give you the tools you need to avoid global hiring headaches.
GROWTH HACK
The "Trojan Horse" Chrome Extension Engine
Turn a free browser tool into a lead-gen machine that captures emails while solving tiny daily annoyances.
The Play: Build a dead-simple Chrome extension that solves one micro-problem for your target user—like a "one-click LinkedIn email finder" or "instant meeting summarizer." Give it away free. Use it to capture emails and subtly surface your main SaaS platform inside the extension interface.
Why This Works:
Chrome extensions live in the browser—the place where your users spend 6+ hours daily. By solving a specific pain point for free, you create habitual usage. Every interaction is a touchpoint. The extension isn't your product; it's your distribution channel. The email capture happens naturally (users expect accounts for synced features), and your upsell surfaces contextually when users hit limitations.
The Implementation Stack:
Category | Tool | Notes |
|---|---|---|
Extension Framework | Plasmo | Free, React-based, hot reload |
Email Capture | Supabase Auth | Free tier: 50K MAUs |
Analytics | PostHog | Free tier: 1M events/month |
CRM Sync | Pipedrive API | Push leads automatically |
Upsell UI | React Modal | In-extension upgrade prompts |
Step 1: Plasmo (The Framework)
Plasmo handles all Chrome extension boilerplate. Initialize with:
pnpm create plasmo --with-tailwindcss
cd my-extension && pnpm dev
Your extension hot-reloads during development. No webpack config nightmares.
Step 2: Supabase (Email Capture)
Add Google OAuth for frictionless signup:
import { createClient } from '@supabase/supabase-js'
const supabase = createClient(SUPABASE_URL, SUPABASE_ANON_KEY)
const signIn = async () => {
const { data, error } = await supabase.auth.signInWithOAuth({
provider: 'google',
options: { redirectTo: chrome.identity.getRedirectURL() }
})
}
Step 3: PostHog (Usage Analytics)
Track which features users love (and where they bounce):
posthog.capture('feature_used', {
feature: 'email_finder',
linkedin_url: url,
user_tier: 'free'
})
Step 4: The Upsell Trigger
When users hit the free limit, show a contextual modal:
if (dailySearches >= 5) {
showModal({
title: "You're on fire today 🔥",
body: "Upgrade to Pro for unlimited searches + CRM sync",
cta: "See Plans",
link: "https://yourapp.com/pricing?ref=extension"
})
}
Step 5: CRM Sync
Push captured leads to Pipedrive automatically:
await fetch('https://api.pipedrive.com/v1/persons', {
method: 'POST',
headers: { 'Authorization': `Bearer ${PIPEDRIVE_TOKEN}` },
body: JSON.stringify({
name: user.name,
email: user.email,
label: 'extension_signup'
})
})
The Result:
CAC: $0 (organic installs via Chrome Web Store)
Email capture rate: 35-50% of active users
Upsell conversion: 3-7% (contextual > cold)
You're not building a side project—you're building a lead pipeline that lives where your users already work.
DAILY STAT
272 Million
Children and youth out of school globally (2025 UNESCO estimate)
The scale: That's roughly the entire population of Indonesia—the world's fourth-largest country—sitting at home instead of in classrooms.
The human comparison: 1 in 10 primary-age children worldwide are denied basic education. In West & Central Africa, it's 1 in 4. In conflict zones, it's 1 in 3.
The Shift:
This isn't improving. The out-of-school population fell by just 1% since the UN Sustainable Development Goals were adopted in 2015—while 110 million more kids entered school. We're running in place. Sub-Saharan Africa now hosts more than half of all out-of-school children globally, and the number there actually increased by 12 million.
The Economics:
High-income countries spend $8,543 per learner. Low- and middle-income countries? $55. That's a 155x gap. African nations now spend almost as much on debt servicing as they do on education. Global education aid dropped from 9.3% to 7.6% between 2019-2022.
What This Means for Builders:
AI-translated education content could reach millions who lack teachers in their native language. The distribution bottleneck isn't content—it's localized delivery.
Micro-monetization models (ads, sponsorships, social good partnerships) could fund free education at scale once volume exists.
Mobile-first is the only path. These populations have phones before laptops, data before wifi.
Gamified learning that works offline and syncs opportunistically could bypass infrastructure gaps entirely.
The opportunity isn't charity—it's building the rails for the next billion learners. Whoever solves distribution to underserved regions captures a market that doesn't exist yet.
TOOL TIP
Blink.new — Ship Production Apps by Chatting With AI
URL: blink.new
What it does: Blink is an AI-powered app builder that generates full-stack applications from natural language descriptions. You describe what you want ("Build me a SaaS dashboard with user auth and Stripe billing"), and Blink writes the code, configures the database, sets up authentication, and deploys—all without you touching a terminal.
Pricing:
Tier | Price | Credits | Best For |
|---|---|---|---|
Free | $0 | 10/month | Exploring, tiny prototypes |
Starter | $25/mo ($21 annual) | 100/month | Solo builders, MVPs |
Pro | $50/mo ($42 annual) | 200/month | Active indie hackers |
Max | $200-$12,500/mo | 800-50,000/month | Teams, agencies |
Who it's for:
Non-technical founders — Go from idea to working MVP without hiring a developer
Product managers — Build internal tools and prototypes to validate before engineering sprint allocation
Indie hackers — Ship faster than competitors who still configure webpack manually
Agencies — Rapid client prototyping with code export for handoff
What makes it different:
True full-stack generation — Database, auth, APIs, and hosting are configured automatically, not just UI scaffolding
180+ model access — Switch between Claude, GPT-4, Gemini, and others mid-project
Built-in agent tools — Web search, code execution sandbox, vector DB, and image generation are native
Code export + GitHub sync — Not locked in; export your codebase and keep building elsewhere
Core capabilities:
Firebase and Supabase integrations that "just work"
Multi-agent workflows with human-in-the-loop approval gates
Self-correction that attempts bug fixes automatically
Real-time preview with hot reload during prompting
Limitations:
Free credits reset daily—heavy experimentation burns quota fast
Generated UI can feel template-ish without design input
Complex apps still benefit from developer refinement post-export
The bottom line: Blink is the fastest path from "I have an idea" to "here's a working link"—especially for founders who'd rather validate with customers than argue with deployment scripts.
TICKER WATCH
Intellia Therapeutics (NASDAQ: NTLA) — $14.81
The Numbers That Matter:
Metric | Value |
|---|---|
Current Price | $14.81 |
52-Week Low | $5.90 |
52-Week High | $28.25 |
Market Cap | $1.72B |
Analyst Target | $4–$106 (avg ~$16-22) |
Cash Runway | $670M (funds into mid-2027) |
What They Do (Simple Version):
Intellia is building a "find and replace" tool for human DNA. They use CRISPR technology—essentially molecular scissors—to edit disease-causing genes directly inside your body. Think: one injection that permanently fixes a genetic disorder, rather than a lifetime of pills.
Their lead programs target hereditary angioedema (HAE, which causes severe swelling attacks) and transthyretin amyloidosis (ATTR, which damages the heart and nerves). Partners include Regeneron and Novartis.
Why This Matters:
The stock is down ~50% from recent highs after a patient death in a Phase 3 trial caused an FDA safety hold on their ATTR program. That's real risk. But the HAE program (lonvo-z) is fully enrolled and on track—with Phase 3 data expected mid-2026 and potential US commercial launch in early 2027.
Three things changed: (1) FDA partially lifted the safety hold on the ATTRv-PN side in January 2026, (2) Cathie Wood's ARK has been accumulating shares aggressively for weeks—adding 40K-57K shares on multiple days, and (3) The company still has $670M in cash, funding operations into mid-2027 without dilution.
The Upside Case:
Conservative: $16-18 target = +8-22% from here
Bull case: $54-106 (some analysts) = +265-615% from here
Profitability: Still pre-revenue; path to profitability depends on HAE approval
Simple Math: $1,000 invested today could become $1,220 (conservative) or $3,650 (bull case) within 12-18 months if clinical data reads out positively.
The Math:
CRISPR gene editing is a "once and done" therapy—one treatment, potentially lifetime results. That pricing power is enormous (competitor Alnylam prices ATTR treatment at ~$450K/year). If Intellia gets HAE approved, peak sales estimates range $2-4B annually. At 5x peak sales multiple, that's a $10-20B market cap (vs. $1.7B today). But binary clinical risk means the math only works if trials succeed.
The Risks (Be Honest):
Patient death in ATTR trial damaged credibility and created FDA uncertainty
Phase 3 data is binary—miss the endpoint and the stock craters
CRISPR is still novel; long-term safety data doesn't exist
Cash runway into 2027 is adequate but not infinite—dilution possible if trials extend
Competition from Beam, Editas, and established pharma partners isn't sleeping
The Verdict:
This is high-risk biotech at a crossroads. The HAE program could deliver a commercial product by 2027—that's the bull case. The bear case is more safety issues, trial delays, and a death spiral to single digits. ARK's aggressive buying suggests institutional conviction, but Cathie Wood has been wrong before.
Position: Speculative. 1-2% max. High risk. Only for those who can stomach 50%+ drawdowns if data disappoints.
Not financial advice. Do your own research.
WORKFLOW
The "Deal Won → Project Kickoff" Automation
Setup time: 15 minutes | Weekly value: 2+ hours saved, zero dropped handoffs
Automatically create project cards when deals close, sync customer data, and notify delivery teams—without anyone touching a spreadsheet.
The Architecture:
Trigger: Deal marked "Won" in Pipedrive
↓
Action 1: Webhook fires → Zapier/n8n catches event
↓
Action 2: Typeform pre-fills → Generate intake form link
↓
Action 3: Trello card created → Populated with deal data
↓
Action 4: Slack notification (if card created)
↓
Outcome: Delivery team has everything needed to start
Step 1: Pipedrive Webhook (The Trigger)
Configure Pipedrive to fire a webhook when deals hit "Won":
{
"event": "updated.deal",
"current": {
"stage_id": 5,
"status": "won",
"title": "Acme Corp - Enterprise Plan",
"person_name": "Jane Smith",
"person_email": "[email protected]",
"value": 24000
}
}
Signal: The webhook payload contains everything the delivery team needs—name, email, deal value, and context.
Step 2: Zapier/n8n (The Router)
Catch the webhook and route data to downstream actions:
// n8n Function Node
const dealData = $input.first().json;
if (dealData.current.status === 'won') {
return {
customer_name: dealData.current.person_name,
customer_email: dealData.current.person_email,
deal_title: dealData.current.title,
deal_value: dealData.current.value,
typeform_prefill: `https://yourform.typeform.com/intake#name=${encodeURIComponent(dealData.current.person_name)}&email=${dealData.current.person_email}`
}
}
Signal: Pre-filled Typeform links reduce customer friction—they click and half the fields are done.
Step 3: Trello Card Creation (The Handoff)
Create a card in your delivery board with all context attached:
{
"name": "Onboard: Acme Corp - Enterprise Plan",
"desc": "**Customer:** Jane Smith\n**Email:** [email protected]\n**Deal Value:** $24,000\n**Intake Form:** [Link]\n**Pipedrive Deal:** [Link]",
"idList": "{{ONBOARDING_LIST_ID}}",
"due": "{{DATE_PLUS_7_DAYS}}"
}
Step 4: Slack Alert (The Notification)
{
"channel": "#delivery-team",
"text": "🎉 New deal closed! *Acme Corp - Enterprise Plan* ($24,000)\nOnboarding card created → <https://trello.com/c/xyz|View in Trello>"
}
Expansion Ideas:
Add Google Calendar event for kickoff call based on Typeform response
Trigger HubSpot sequence for onboarding email drip
Create Notion database entry for project tracking alongside Trello
Auto-generate SOW draft in Google Docs using deal details
Sales closes the deal. Delivery starts the project. Nobody asks "where's the customer info?" ever again.
THE BOTTOM LINE
Claude's Desktop Extensions aren't just a convenience feature—they're Anthropic's bid to make AI assistants useful for people who don't live in terminals. When the friction between "I wish Claude could access my files" and "Claude can access my files" drops to a single click, adoption curves bend upward.
Connect that to the education stat: 272 million kids can't access classrooms, but most have access to a phone. The distribution problem isn't content creation—we can generate localized, translated, gamified education with AI today. The problem is getting it to devices in underserved regions in ways that work offline, monetize sustainably, and don't require infrastructure that doesn't exist.
The playbook: Build where friction is disappearing fastest. Desktop Extensions mean non-technical users can now leverage MCP integrations—build tools that take advantage of that expanded market. The Chrome Extension growth hack works because it meets users where they already are, captures emails through natural value exchange, and surfaces upsells contextually. The Pipedrive-Trello workflow eliminates handoff failures that kill customer experience post-sale.
If you're building, ask: "What's the smallest useful thing I can ship that creates a daily habit?" If you're investing, biotech like NTLA offers asymmetric upside but demands position sizing discipline—binary events destroy undiversified portfolios. If you're selling, automate the boring handoffs so your team focuses on the work that actually requires human judgment.
Ship daily.
HackrLife Daily is read by growth marketers at Google, Adobe, LinkedIn, and creators building the future.