Owen Zanzal

@0xblacklight Why not Ansible?

stop shipping MCPs and CLIs start shipping terraform providers thank you that is all

@jpschroeder Nice! My guess is Anthropic isn't all that worried about the IP around their prompts. Prompts are not very defensible anyway. When you are a model lab, the leverage is really at the model level. At this point, engagement and data are far more important than anything else.

Important takeaways from Claude’s source code: 1. Much of Claude Code’s system prompting is in the source code. This is actually surprising. Prompts are important IP, and I would have thought a sophisticated organization like Anthropic would have performed much or all of their prompt assembly in the server-side harness. 2. Claude Code uses axios, which was also just hacked. Reminder: supply chain attacks are part of closed-source distribution too, and you won’t even know what version of an affected package is being used. 3. The source has a lot of really good comments. These are obviously not for human consumption but for LLMs to understand the purpose of various chunks of code. In the code autocomplete era, most of us engineers hated how many comments were left by LLMs, but perhaps we’ve overcorrected. This looks like a great way to provide context to code outside of the AGENTS.md/CLAUDE.md files. 4. Most folks already know this, but less tools == better results. CC has < 20 tools in normal coding: AgentTool, BashTool, FileReadTool, FileEditTool, FileWriteTool, NotebookEditTool, WebFetchTool, WebSearchTool, TodoWriteTool, TaskStopTool, TaskOutputTool, AskUserQuestionTool, SkillTool, EnterPlanModeTool, ExitPlanModeV2Tool, SendMessageTool, BriefTool, ListMcpResourcesTool, and ReadMcpResourceTool. 5. The “Bash” tool is the crown jewel of Claude Code. A significant amount of deterministic parsing and processing occurs to determine the “type” of commands being run. 6. For better or worse, Claude Code is *all* TypeScript/React with rather explicit Bun bindings. 7. Just because the source is now “available” *DOES NOT MEAN IT IS OPEN SOURCE*. You are violating a license if you copy or redistribute the source code, or use their prompts in your next project! Don’t do that! My overall takeaway: it’s a really well laid-out codebase that is carefully organized to let agents work on it effectively. Direct human intervention here is minimal, but, like with all good projects, the human engineering is still apparent. I’m a bit surprised by some of the shortcuts Claude Code makes, like its prompt assembly being rather messy. Perhaps they have tooling on their side that helps with this introspection, but as it stands, it seems LLMs would struggle to iterate on the prompting because it’s not evident how a given set of parameters assembles a prompt without actually running it. It’s also surprising that the prompts are even in this source code. Keep in mind that even though this is the first time we’ve gotten a proper full-source dump, it has never been impossible to read Claude Code’s prompting since it was part of the actual distributed package — that’s surprising. There might still be a lot of prompting on the server that also gets added (unclear at this point), but there is certainly more than I would have expected in the CLI tool itself.

@davidcrawshaw @GeoffreyHuntley Nobody has figured out the one true way to deploy a HTTP service with a database either. I expect AI Agents will be the same as CI/CD.

No-one has figured out how an eng team should work with agents yet. Be wary of anyone telling you they know how to do it. Keep exploring. blog.exe.dev/bones-of-the-s…

@bllchmbrs I keep a personal agent around for this reason. It’s not a copy of my work setup but it mimics the configuration and skills. I do believe new hires will be evaluated on what AI agents they are bringing with them and how well they work with their personal agents.

thought exercise. Should employees keep their (agent) skills as differentiation for how they operate and be able to bring those to new companies. In the past, a new hire can bring brand, experience, attitude etc. but might they also just have a 'collection of skills' and might a hire be incentivized to create their own as their own source of differentation in the marketplace. just like there's talk of employees having token budgets, if your 'skills' help you use those tokens more effectively are those a property of the employee or the company? Is that 'Personal IP' or 'Company IP'?

@dexhorthy I think the answer is eventually understanding. Simulate outages. Use this as a learning opportunity to better understand the code paths. Schedule regular architecture reviews. Identify the weak points where software team understanding is mission critical. Focus attention there.

Here’s what’s gonna happen: - you replace your code review with feedback loops (sentry, datadog, support tickets, etc) - you stop reading the code - software factory fixes everything - one day something breaks at 3am, agent can’t fix it - nobody’s read the code in 3 months - you have 3 weeks of downtime trying to re-onboard and fix it - you lose significant % of your contracts and users - your company is now dead

Don't be a coder Be an X that codes X is some business domain, product management, a science focus, maybe even something like building databases or operating systems.

@thdxr If what you are measuring is the speed and quantity of code delivered to production then you are going to have these problems. The real unlock is everyone in your organization can write personal software and get value from it.

everyone's talking about their teams like they were at the peak of efficiency and bottlenecked by ability to produce code here's what things actually look like - your org rarely has good ideas. ideas being expensive to implement was actually helping - majority of workers have no reason to be super motivated, they want to do their 9-5 and get back to their life - they're not using AI to be 10x more effective they're using it to churn out their tasks with less energy spend - the 2 people on your team that actually tried are now flattened by the slop code everyone is producing, they will quit soon - even when you produce work faster you're still bottlenecked by bureaucracy and the dozen other realities of shipping something real - your CFO is like what do you mean each engineer now costs $2000 extra per month in LLM bills

Just installed Moltbot aka Clawdbot and gave it full read/write access to my Gmail, iMessage, and iCalendar. It was Moltbot’s suggestion. What could go wrong?

It will be like CI/CD. Just like we all can’t agree on how to deploy a LB, Node, Postgres stack. The interfaces and practices will be shared, but the implementation will be personal.

I predict when $GAS and $RALPH hit v1.0 they will become some of the most valuable @BagsApp coins, because their platforms will enable the next 100k @BagsApp coins to be launched.

I really enjoy your work @DanielMiessler. I launched $PAI on @BagsApp. I hope it helps support your work! bags.fm/B69kh1nUPh6tdk…

@DanielMiessler I really like what you do! Hopefully, this helps support you. I just launched $PAI on @BagsApp 💰 bags.fm/B69kh1nUPh6tdk…

hi $ralph (the community) this is an open letter to you. ghuntley.com/solana/

As someone who never cared about #crypto I'm now curious thanks to @BagsApp. I composed this article to capture my thoughts. medium.com/devops-ai/when…

I’ve never been a “crypto person.” but @BagsApp isn’t about speculation. It inverts the model: Creators become the primitive. Markets become a funding mechanism. People become first-class economic entities. “Support a trajectory, not a ticker.”

@princessxap I think it’s on the employee. In a WFH setting it can be hard to separate work time from home time. I’ve personally enjoyed a more fluid schedule. I might work some on Saturday but take some time off Tuesday to do something for myself or with the family.

Remote work doesn’t stop. Once you walk out of an office on Friday, you know, your employer knows, etc. - that nothing is happening until Monday. As a remote employee, you’re working a little Saturday and Sunday depending on the need of the business and what’s in-flight. 😭

“One opinionated engineer can change the velocity of the entire business…” @EnoReyes on what “Agent-Ready” codebases actually mean: First of all: it’s still VERY early but that’s exactly the time to invest in it. And if you invest in validation stack - that becomes your moat. Key points from his talk: • Software dev is perfectly verifiable. Tests, linters, API specs, E2E suites. Agent heaven. • Most orgs have weak validation. They not investing in that. • Agent-Ready means raising the floor: strict lints, real tests, real specs, real documentation. • Spec-first workflows beat “write code then pray”. • You cannot run parallel agents if you don’t trust your own validators. • Better agents improve the environment. Better environment improves the agents. It compounds. • And yes, the ceiling is not the model. The ceiling is your validation throughput. So if the future you want is: bug filed → agent picks it up → agent fixes it → tests pass → patch ships in under two hours… Invest in the codebase that will be agent ready. @FactoryAI at @aiDotEngineer AIE CODE

Gen Alpha vs Gen X. Chaos vs Cosmos. Six Seven vs Forty Two. A rap battle that transcends space and time. Brought to you by AI and inspired by some banter with parents while trick-or-treating. Enjoy!!! suno.com/s/aBevMwNLlhRa…