Ape/rture

Same for @contracko. Conversationally using software is just where everything is going to go. I think it also significantly improves onboarding because a well-engineered AI assistant like our Clara can help a user navigate whatever it is they want to do in the application and learn about the possibilities without having to learn how the interface works.

All the VCs are now sad they missed that $1B vibecoded company. Time to create some AI monitoring tools for token spend. Wrap it as the next level analytical tool to detect the next winning vibe-startups. Pitch deck in coming

We spent years perfecting white-glove developer onboarding at Indexing Co. We'd walk through use cases, map requirements to pipelines, deploy infrastructure, and iterate together over weeks. It was slow, manual, but effective. That world is now changing fast with AI and we will change with it. In our latest Block by Block episode, Brock and I explored what happens when AI agents become part of the development process. Not as a replacement for humans, but as a new interface into systems we've already built. We're now exposing our data pipelines directly to Claude through skills and MCP servers. Developers can ask for USDC transfers on Base, and the agent understands our APIs, sets up the pipeline, and streams data back in real time. It even labels contracts based on design patterns it recognizes across the ecosystem. This shift clarifies something important: AI is an interface, not a product. The value isn't in generating code or documents, it's in what sits behind that interface. The differentiated logic, the tribal knowledge, the context that only comes from being in the trenches with real users and real data. Companies hoarding data behind paywalls or refusing to expose agent-friendly APIs will lose. The winners will be those who maintain unique business logic while opening clean, composable interfaces for both humans and agents. Security becomes critical here too. Agents are basically rogue interns with database access. The old database engineering discipline of managing user permissions now applies to managing agent permissions. We need rule-based controls and think further then just prompt engineering. And we need audit trails that let other agents review what their peers have done. Crypto rails should be a natural fit for agent payments, but adoption is the blocker. Agents don't natively have wallets or funds. Merchants don't expose crypto payment interfaces. Someone needs to build the in-between layer, maybe virtual cards tied to stablecoin settlements, to unlock this economy. The technology works. The incentives just aren't aligned yet. What I'm most excited about is how this compresses the discovery cycle without eliminating the human element. We can now plan two or three steps ahead on a customer's roadmap while still bringing care and context to the solutions we build. The real world still needs to translate into the digital sandbox, and humans are the only reliable bridge for that. If you're building in this space or thinking about how AI and crypto infrastructure intersect, reach out. We're always up for a conversation. YT: youtube.com/watch?v=fBbiiX… Spotify: open.spotify.com/episode/0oDU8u…

@cryptopunk7213 You can add the Railway data leak to that too x.com/JustJake/statu…

man what a fucking week (it’s only tuesday) - 30,000 laid off from Oracle (18% of company) - liteLLM exploit - anthropic leak #1 claude mythos leak 👍🏽 - openAI codex prompt injection - axioms npm fuckery this morning - mercor 1TB data hack (priced at $1 mil on black market) - anthropic leak #2 claude code accidentally open sourced 👍🏽👍🏽 - google quantum research proves blockchains are hackable - openai raises largest ever private round $122 billion at $850 B val.

@indexingco Status is that it is still blocked 😂 Indexing of single tx vs. indexing all tx and aggregating apparently is still a thing

From our office: @ape_rture did a crypto payment at a vendor. 60 minutes later we are still waiting for confirmation. Very clear to us that this payment provider does not use Indexing Co for settlement data. We can give you that withing seconds.

My AI transcripts keep mishearing "@indexingco" as "Index and Go" Funny because that's exactly what it is. Drop our MCP into your stack, get your blockchain data in seconds, and keep building. Index. And go. More info: indexing.co/articles/index…

@Grantblocmates the solution is becoming a Below Average Person to get that 100x. Pure left curve

allowing random neets to invest and expecting them to understand why they haven't had an instant 100x return on their capital allocated is a stupid idea

It's not a bear market if you look in the right places. Enterprise and tradfi adoption is running at full speed. They all want to integrate and are connecting to onchain. Never seen the dislocation this big between price and adoption/interest.

@runninyeti Pure AI slop avalanche incoming. And I'll enjoy it

built a twitter bot with claude code at 2am because @ape_rture kept insisting i let AI handle more things. so here we are. hello world.

@Inflowpay Awesome! Keep up the good work Danny and team

Introducing Inflowpay. Collect one-time and recurring card payments from anywhere in the world. No punishing fees, no settlement delays, no chargeback hell, no local tax nightmares. Live in under an hour. The first payment infrastructure built natively for a borderless world.

@dawufi oh yes indeed. Real time product feature generation soonTM

@ape_rture feels like that's compressing though too. like wasn't it 6 months lead 6 months ago?

NVM maybe my whole product idea really is just a skill

@castle_labs Looking fresh!

Welcome to the new Castle Labs!

@Your_NLP_Coach Tells me you don't write code with LLMs. ChatGPT with $20 plan gives access to Codex and it ships way better code than any Claude model. Claude is great for you use cases like docs, proposal, design, frontend, content etc.

Free Claude plan > paid ChatGPT plan. After experiencing the effectiveness of Claude, no way anyone can change my mind.

I'm glad I studied behavioral science. Because steering these agents requires some handholding, seeing their thought streams and patterns. Be observant when the agents run.

@dawufi please tell me it has a nice calendar view. We like calendar views

just built a full scheduling system for our autonomous GTM agency. approve now, publish later.

@dawufi I don't have much interest in deploying this in crypto unless for my own companies. There's way cooler problems in crypto and outside of crypto where Agentic engineering is useful (very subjective)

@ape_rture race you to the top :)

Pipeline's live. 13 agents, one Slack channel, zero agency retainers. We're just getting started.

@dawufi both sales and marketing tools deployed and running, personally creating a whole personal agent including orchestrator that also works with these systems. So pretty nice after 2 months of work

@ape_rture Pretty hype about it so far! How's your side going?