How do you get Claude Code to check its own work before handing it back? Watch how you can encode your manual checks so Claude closes its own feedback loop:

Token costs are why there will be no saas apocalypse / good dev tools are cached intelligence for agents! The popular theory goes: agents can write code, so they'll just rebuild every tool from scratch and hit raw APIs. no more dev tools, no more CLIs, no more software layers. just agents and endpoints! We just tested this and the data says the opposite. We benchmarked Claude Code and Codex on real Hugging Face Hub tasks (~1,000 graded runs), with two setups: the agent-optimized hf CLI vs the agent hand-rolling curl or SDK calls from scratch. Hand-rolling burns up to 6x more tokens on multi-step tasks and fails more often (84% vs 94% task success). And that's just dropping one abstraction layer. It would obviously be orders of magnitude more tokens and a dramatically higher failure rate if the agent tried to bypass HF altogether and rebuild model hosting, versioning, and distribution from scratch. Every time an agent re-derives a workflow from raw API calls, you pay for that reasoning in tokens. every single run. a good CLI compresses that entire chain into a few high-level commands the agent can't get wrong. In a world where everyone is complaining tokens are too expensive, abstraction is leverage: thousands of hours of design decisions your agent doesn't have to re-reason about at inference time. Good tools are cached intelligence for agents! So no, agents won't rebuild everything from scratch. they'll gravitate to the most token-efficient tools, because that's what their owners pay for. The software that survives won't just be accessible to agents, it will be accurate and cheap for them to drive. We're seeing it happen with HF, which is becoming the platform for agents to use AI: ~49M requests in just two months, and growing fast! huggingface.co/blog/hf-cli-f…

AI should earn its keep. Introducing the AI Productivity Guarantee. If Devin delivers less engineering value than you’re paying for, Cognition will fund your usage until it does, up to $10 million. It’s time for the AI industry to stop maximizing tokens and start maximizing productive output.

Automatically mine and create agents for any business. Over the weekend I went to the Google IO hackathon and had a blast building what I called Mitosis. Generates a synthetic company to create conversations and people, automatically extract the workflows from the company based on the conversations and build out new agents to automate the busy work. Pushes all context into the company brain, leverages governance and policy gating to make sure the right person stays in the loop and the agents learn from their mistakes as they make them to get better over time. The end result is a 95% drop off of handoff between the employees in this example. Each week the agents can take on more to empower the employee to get even more work done while making the important judgement calls. This is all of course synthetic and the milege will vary greatly per company and workflows but was a great experiment to build in a day and leverage a lot of the work I have been working on for other businesses. It was also great to try out the antigravity managed agents and Gemini 3.5 Flash. In this demo you may see that it is listed as mock, this is definitely mocked. In the hackathon we had an account with free credits. I burned through the rate limits and the credits rather quickly. The rate limit for the anitgravity agents was 200k a minute, I blew past it at 2M a minute. So this was definitely tokenmaxxing. I look forward to expanding this further to other platforms and technology partners but was a blast to build.

Introducing Claude Opus 4.8: it builds on Opus 4.7 with sharper judgment, more honesty about its own progress, and the ability to work independently for longer than its predecessors. Available today at the same price.

Over at Substack, @JoshEakle asks: "It's 2026, and I have yet to see an anti-almond farm protest."

The hottest job for the next five years is going to be the agent operator. They don't need to be an engineer. They can walk into marketing, legal, or life sciences research and actually make agents work for that function. Required skills: > MCPs > CLIs > Writing skills (the file kind) > agents.md fluency > Business acumen None of this is in any CS curriculum today. Soon, enterprises will be pressured to redesign their workflows for agents, not for people. And when that happens, agent operators will be in massive demand.