26 Articles
What Matters When You Build a Machine for Local LLMs
Five hardware factors, ranked. Capacity decides whether it runs at all. Bandwidth decides how fast it feels.
Fidelity is up and obedience is down!
As a developer, you will be silently fired if you execute on user stories exactly as written. It won't be on purpose and it will happen slowly over time. This is called "ticket obedience" and it is no longer rewarded in the age of AI.
USB cables look the same. They are not.
Most of what matters on a USB cable is invisible. We organized the mess into a simple reference at usb.periscoped.io.
Announcing ontology.periscoped.io: A Worldwide Index of Prior Art for Your Data Model
A listing of major and ontologies and other related research.
The Ontology Already Exists: How to Find It Before You Build One
Someone has probably already modeled your domain. A field guide to the world's ontology registries — and how to reuse and enhance the right model fast.
Build an Ontology: The Competitive Advantage Hiding in Your Data
Your data isn't the problem, your model is. Here's how building an ontology turns a pile of data into a competitive advantage your AI and team can trust.
Stop Building on Shifting Sand: Lock Your Object Model Before AI Touches It
AI-assisted projects turn to slop when there's no foundation. Lock two files — a plain-English object definition and a derived data model — and AI has to build on rock instead of shifting sand.
We Gave Every AI Agent a Job Title: It Changed Everything
We broke AI agents into domain-specific subagents with explicit ownership of routes, types, and business rules. The codebase got better overnight.
AWS Identity Center: How to Stop Giving Everyone the Keys to the Kingdom
Managing AWS permissions at scale is a losing game — everyone ends up with admin access. Here's how structured review sessions and an automated reporting agent can bring your Identity Center back under control.
Claude Hierarchy
Stop Claude Code from interrupting you every few seconds. Get the settings hierarchy right — and remember the rule that matters most: your local settings always win.
Your AI Tooling Needs Maintenance Too
The .claude/ directory drifts just like production code. Here's how a dedicated folder manager agent keeps your AI tooling configuration clean and working.
The API Agent That Enforces What Developers Forget Under Pressure
Under deadline pressure, API routes accumulate bad habits. Here's how a dedicated agent enforces REST standards, Zod validation, and permission checks on every route, every time.
Giving the Frontend Its Own Lane
The frontend is always last and always rushed. Here's how a dedicated frontend agent with a fixed execution order eliminated merge conflicts and built UI against stable contracts.
We Built an Agent That Runs Our Entire Sprint While We Sleep
Sprint execution is project management with a keyboard. Here's how a sprint-runner agent handles every ticket end-to-end — no human in the loop.
Why We Put Every Database Migration Behind Its Own Agent
Schema changes are where features go wrong. Here's how isolating database work behind a dedicated agent eliminated a whole class of deployment failures.
We Stopped Writing Tests After the Fact. Here's What We Do Instead.
Tests written after a feature ships are just documentation. Here's how we made testing parallel with implementation using a dedicated test agent.
A Well-Written Ticket Is Half the Implementation
The ticket says 'add search.' That's it. Here's how a structured ticket agent eliminates vague requirements before a line of code gets written.
The Smallest Problem That Wastes the Most Time
Fifteen minutes of friction spread across a hundred small moments. Here's how a design asset agent eliminated the constant overhead of placing files in the right place.
How We Built an Agent That Turns Jira Tickets Into Shipped Features
Features miss deadlines not because engineers are slow, but because coordination is invisible overhead. Here's how we built an agent that handles the orchestration so humans don't have to.
We Built an AI Skill That Reviews Our Case Studies in Seconds
Case studies are high-value but slow to produce. We built an AI-powered review skill that checks completeness, flags confidentiality concerns, validates SEO, and gives authors a clear checklist — turning weeks of review into seconds.
Building The Best CLAUDE.md File
Your CLAUDE.md file is the brain of your AI-assisted project — neglect it and your AI starts working against you.
MCP Servers: The Hidden Productivity Killer No One Talks About
MCP server connections break constantly, killing developer flow. Here is how to configure them at the project level so your team stops losing hours to connection issues.
Be lazy for five minutes
Nothing makes you want to work more than not being allowed to work.
Micro Improvements: Your Engineering Edge
Complaining is for losers. The best engineers shut up and grind — improving everything around them, bit by bit, without fanfare.
Never Develop Without Seeded Data
Not having fully seeded test data for your user story demos and acceptance criteria are a disaster waiting to happen. Just build it first, and worry about the data last. That's actually the wrong approach in LLM world
Tiny Details Matter - Field Level Focus
It all starts at the smallest atomic level, which is your database fields. In order to build great software, you need to know the smallest details to include architecture, fields types, and field restrictions on a single data point.