Manage Your Agent’s Loadout with Dr. Skill
Keep your context clean by keeping your agent's skills and tools in check.
Drew Breunig
Writing about AI, geo, culture, media, data, and the ways they interact.
Latest posts
Understanding the Dynamics of the AI Ecosystem with Pace Layers
Stewart Brand's Pace Layers framework reveals why AI's speed is straining the slower systems it depends on for support.
The Problem is Prompt Debt
The plain-English prompt that makes prototypes effortless turns out to be a poor way to specify how a system should behave, and the bill arrives slowly, disguised as ordinary progress, until the application can barely move.
What Do Humans Need From Docs?
When an agent can explain anything, should we keep writing docs for humans?
The Cost of Overfitting the Harness
OpenAI winding down fine tuning is an interesting development and one to watch.
10 Lessons for Agentic Coding
Guidelines for programming when code is cheap.
Cybersecurity Looks Like Proof of Work Now
Is security spending more tokens than your attacker?
How Claude Code Builds a System Prompt
Context engineering in action.
The 2nd Phase of Agentic Development
Moving from clones to reimaginings.
The Cathedral, the Bazaar, and the Winchester Mystery House
Welcome to the era of sprawling, idiosyncratic tooling.
Learnings from a No-Code Library: Keeping the Spec Driven Development Triangle in Sync
The following is a write up of a talk I delivered at MLOps Community’s “Coding Agents” conference, on March 3rd. There’s a video version of the talk available on YouTube.
We’re Talking About Terms of Use, But the Issue is Embedded Judgment
The biggest buyers will want to audit and influence post-training
Two Beliefs About Coding Agents
There’s a lot of noise about how AI is changing programming these days. It can be a bit overwhelming.
Why is Claude an Electron App?
If code is free, why aren’t all apps native?
How System Prompts Define Agent Behavior
System prompts matter far more than most assume. A given model sets the theoretical ceiling of an agent’s performance, but the system prompt determines whether this peak is reached.
The Potential of RLMs
Handling Your Long Context Today & Designing Your Agent Tomorrow
The Rise of Spec Driven Development
It’s been a month since I launched whenwords , and since then there’s been a flurry of experiments with spec driven development (SDD): using coding agents to implement software using only a detailed text spec and a collection of conformance tests . Github Could Use a ‘Docs Review’ UI First off, despite whenwords being a couple Markdown docs and a YAML test set, people have submitted valuable PRs.…
A Software Library with No Code
Do we still need libraries of 3rd party code when AI agents are this good?
2025 in Review: Jagged Intelligence Becomes a Fault Line
Looking back on 2025, the incredible pace of AI is stunning. But fast growth brings disconnects.
Why I Write (And You Should Too!)
Publicly writing is one of the most valuable things you can do. It improves your thinking, ability to communicate, and connects you to interesting people.