Your AI Coding Agent Is Not Sandboxed Just Because It Runs in DockerLearn why Docker alone cannot secure AI coding agents against hostile repositories, VM isolation, credential separation, Git hardening reduce risk.18 min read·basiliskin.hashnode.devOct 12
Why CQRS Is a Surprisingly Good Fit for Orchestrating Coding AgentsAgents are long-running processes that need human input and recovery. Here’s how CQRS turn agent orchestration into a resilient, testable workflow.13 min read·basiliskin.hashnode.dev3AOct 6
Feature Flags Without a Vendor: Why I Replaced PostHog With 200 Lines of S3Every feature flag SDK sells you the same promise: flip a switch, change behavior in production, no deploy. What the pricing page doesn't show is the second contract you sign — your request path now d22 min read·basiliskin.hashnode.devSep 22
Event-Sourced AI Agents: Building Frontier-Quality Coding Agents on 32k Context WindowsHow to build autonomous coding agents that run 24/7 on small local models — without losing quality, without breaking the loop, and without burning through tokens. 1. The Problem Everyone Ignores Moder21 min read·basiliskin.hashnode.devSep 15
The Missing Observability Layer for Your LLM Command-Line Tools: Introducing cli-obs-proxyIf you’re a developer who loves working in the terminal, you’ve probably adopted command-line LLM tools to speed up your workflow. But if you’re like me, you don’t just stick to one provider. You jugg5 min read·basiliskin.hashnode.dev2RSep 14
