! 22 articles published
A vendor-agnostic, YANG-model-driven MCP service that enables LLMs to safely interact with network infrastructure — with human operators always in control through structured approval flows and full auditability.
AI agents are starting to look like operating systems. The LLM is the CPU, the agent is the OS, and skills and MCPs are the applications — here's why that analogy holds up.
AI can generate Cisco configs and Terraform plans with impressive fluency, but the hard part of network engineering was never the syntax. It's interop, requirements, and the mental models we build to design and troubleshoot.
Moving beyond vibe coding to structured AI-assisted development — using a .ai/ directory and project constitution to give AI assistants the context they need to be genuinely useful.
Using netconan to strip passwords, IPs, and sensitive strings from Cisco and Juniper configs before sharing them — deterministic, reversible, and prefix-preserving.
Three ways to turn any web app into a standalone macOS application — from the built-in Safari web app feature to lightweight Rust-based wrappers with Pake.
A practical walkthrough of SSH key generation, distribution, agent management, and config file patterns — updated for modern best practices with Ed25519.
Using Netconan and Python to strip passwords, IPs, and sensitive data from Cisco configs before sharing them.
Deploying K3s on a dedicated server with k3sup in minutes — from bare metal to a working cluster, plus the practical next steps to make it useful.
A system prompt for getting clean, production-grade Python from Claude Opus 4.6 and Sonnet 4.5 — with rules for type hints, error handling, design patterns, and code structure.
A practitioner's map of the network automation ecosystem — from CLI parsing and config management to source of truth platforms, testing, and AI-assisted operations.
NetDevOps is not just for operators writing scripts. Network designers need to evolve too — embracing programmability, data models, and infrastructure as code to stay relevant.
How to mitigate gosu security concerns by recompiling it with a current Go toolchain, plus a multi-stage Docker build pattern to pull the binary into your own images.
As LLM accuracy improves from broken syntax to near-flawless output, the real question shifts: how much hard skill do we actually need, and how do we keep it sharp?
Stop manually sourcing env files and juggling export statements. Use direnv and modern secret managers to load project-specific environment variables automatically.
An AI agent prompt that transforms raw input into atomic, Zettelkasten-style permanent notes inside Obsidian — with consistent structure, technical writing standards, and automatic link suggestions.
Applying network engineering discipline to cloud networking — using Terraform to manage Transit Gateway route tables, attachments, and peering with the same rigor as an MPLS backbone.
Building a lightweight BGP route leak detection tool using pyBGPStream and RPKI validation to catch misconfigurations before they propagate.
Using large language models to extract structured data from show command output when TextFSM templates don't exist or are too rigid.
Connecting Nornir to NetBox for dynamic inventory management, so your automation always reflects the actual state of your network.
Using Batfish to validate network configurations before deployment — catching routing loops, unreachable subnets, and policy violations without touching a live device.
Moving from manual ACL management to declarative infrastructure as code using Terraform and Python, eliminating the risk of fat-finger mistakes at 2am.