Joshua Renfro

(👾,👾)

[Jeddi @antinertia]

i advise 10+ companies doing 7 to 9-fig arr

they all ask me the same question:

“who’s the best gtm/growth person we should hire?”

and i have 0 names to give

there’s a massive shortage of elite gtm/growth talent

you can see it in the market

salaries have doubled in 3 years

if you’re a GREAT gtm/growth operator, dm me

i probably have a job for you

[Joshua Renfro @renfrodotdev]

@chrispisarski I do this for a living at @CadreAI

[Joshua Renfro @renfrodotdev]

@J_mili Vercel deployment, verbal feedback

[Joshua Renfro @renfrodotdev]

@unusual_whales Anddddd it’s gone

[Joshua Renfro @renfrodotdev]

@J_mili I had all the tokens 😂
Should’ve just started w your karpathy remix repo. Speaking of which, my team is likely building on for GTM context harnesses (for client work). Would love your thoughts on it in a few weeks.

[Joshua Renfro @renfrodotdev]

@J_mili @karpathy John w the alpha drop. Epic repo shares x2. Best concept I’ve seen in a while

[Claude @claudeai]

You can now enable Claude to use your computer to complete tasks.

It opens your apps, navigates your browser, fills in spreadsheets—anything you'd do sitting at your desk.

Research preview in Claude Cowork and Claude Code, macOS only.

[Corey Ganim @coreyganim]

Becoming a Claude Architect can sound overwhelming so here's the step-by-step checklist:

Week 1: Agentic Loops (Domain 1)
□ Read Agent SDK Overview docs
□ Build a multi-tool agent with 3-4 MCP tools
□ Implement proper stop_reason handling
□ Add a PostToolUse hook to normalize data formats
□ Add a tool call interception hook to block policy violations

Week 2: Tool Design (Domain 2)
□ Write tool descriptions that actually differentiate similar tools
□ Create structured error responses with all 4 categories (transient, validation, business, permission)
□ Configure .mcp.json with environment variable expansion
□ Test tool_choice forced selection

Week 3: Claude Code Config (Domain 3)
□ Set up CLAUDE.md hierarchy (user + project + directory level)
□ Create .claude/rules/ with glob patterns for test files
□ Build a custom skill with context: fork
□ Test plan mode on multi-file refactor vs direct execution on single bug fix

Week 4: Prompt Engineering (Domain 4)
□ Build an extraction tool with JSON schema (required, optional, nullable fields)
□ Implement validation-retry loop
□ Run a batch through the Batches API
□ Add 2-4 few-shot examples for ambiguous cases

Week 5: Context Management (Domain 5)
□ Build coordinator + 2 subagents
□ Implement persistent "case facts" block
□ Simulate timeout with structured error propagation
□ Test with conflicting sources, verify attribution survives

Hopefully this simplifies the process and makes it more approachable.

[vibhu @vibhu]

Email for your Solana clanker is here

[Thariq @trq212]

https://t.co/nKTDfC7zMm

[Joshua Renfro @renfrodotdev]

@amirmxt @HarryStebbings Spot on. But how find anon

[Manthan Gupta @manthanguptaa]

https://t.co/CzmfXDgPRP

[Min Choi @minchoi]

8. Injecting malware into code with Clawdbot
https://t.co/YLUIOUKcVt

[Joshua Renfro @renfrodotdev]

@sidbid Great to see some intersectionality between crypto and AI culture on here

[Sisyphus Labs @justsisyphus]

This is the first useful article that covers agent orchestration

[Joshua Renfro @renfrodotdev]

@ErikSchluntz Rad. Could you program a reachy mini to do the same?

[Boris Cherny @bcherny]

YES

[a16z crypto @a16zcrypto]

https://t.co/oN4Vl9VuXC

[Claude @claudeai]

Introducing Cowork: Claude Code for the rest of your work.

Cowork lets you complete non-technical tasks much like how developers use Claude Code.

[Boris Cherny @bcherny]

I'm Boris and I created Claude Code. Lots of people have asked how I use Claude Code, so I wanted to show off my setup a bit.

My setup might be surprisingly vanilla! Claude Code works great out of the box, so I personally don't customize it much. There is no one correct way to use Claude Code: we intentionally build it in a way that you can use it, customize it, and hack it however you like. Each person on the Claude Code team uses it very differently.

So, here goes.