The first step in the AI workflow is specification. This article shows how I turn a vague user story into something a human and an agent can use without guessing, with Gherkin, typed examples, and properties.
AI
A Markdown knowledge base built to last several years, with an AI agent and vector search. Seven design choices are documented along with their costs.
The implicit promise of AI coding assistants was that tests would become a thing of the past. The reality documented by Kent Beck, ThoughtWorks, and several 2025 studies is the opposite: with AI, tests become essential — but the work has shifted.
pi in YOLO mode gives you full filesystem access, unrestricted command execution, zero guardrails. The creator made that choice deliberately. But when your project contains .env files, SSH keys, or a production.yaml, that choice puts you at risk.
I covered pi-secured-setup a few days ago. Guards, Scanners, audit trail, wired into the agent. Since then I looked at another extension: pi-permission-system by MasuRii. Both secure pi. Not the same way.
In the previous article, I introduced pi-secured-setup — a pi extension that adds Guards, Scanners, and an audit trail to your AI coding agent. It ships with sensible defaults: boundary enforcement, protected path globbing, bash command classification, secret redaction, skill verification.
But every project has unique risks. A Terraform shop needs different rules than a Node.js monorepo. A team with strict compliance requirements needs different audit granularity than a solo developer.
A few days ago, I covered Greywall — a kernel-level sandbox that contains pi with a deny-by-default approach. That’s your outer wall. But what about threats inside the boundary? The agent that accidentally writes to the wrong project, the .env file that ends up in the LLM context, the skill whose SKILL.md was silently modified. That’s a different problem, and it needs a different tool.
Today I’m releasing pi-secured-setup — a pi extension that adds Guards, Scanners, and an audit trail directly inside the agent. No kernel modules, no containers, no external dependencies. Just a pi install and you’re protected.
AI coding agents like pi have become essential daily companions. But by default, pi runs in YOLO mode: full filesystem access, unrestricted command execution, no permissions. It’s a deliberate design choice by its creator, but this freedom comes with real risks. Today, let’s explore Greywall, a tool that sandboxes pi using a deny-by-default approach at the kernel level.
Why Sandbox an AI Coding Agent?
pi in YOLO mode is convenient but risky. Without restrictions, the agent can: