rshade

rshade

Richard Shade · KI5WWE

Notes on infrastructure, FinOps, AI tooling, and the occasional brisket.

topics

notes

a linux mint mini pc for an aging parent →

A Linux Mint mini PC built around remote support, linux for our parents

issues are prompts, PRs are context, github is the queue →

GitHub was always a queue. Issues are the prompts, PRs are the context; swap the human consumer for an agent and the shape doesn't change. It's fine that it's not RabbitMQ.

#ai#projects

gh-aw-fleet 0.2: cost and risk, fleet-wide →

The consumption rollup promised at launch shipped with a --budget flag, and the same fleet-wide view caught Renovate and Dependabot quietly fighting gh-aw's own pins.

#projects#agents#ai#finops

ax-go 0.2 and 0.3: contracts worth pinning →

v0.2.0 split ax-go into contract packages you can import without the runtime; v0.3.0 added the coverage floors and compatibility matrix that keep them safe to pin.

#projects#go#agents#ai

agentic experience for Go →

ax-go is a Go library that makes CLIs predictable for LLM agents: trace IDs that cross the plugin boundary, one function behind the human table and the agent's JSON, and deterministic output.

#projects#agents#ai#go

where the copilot credits went →

Usage-based Copilot billing meters in AI credits now. The per-repo tools can't see across a fleet. One fleet.json rolls the spend up and finds a single agent eating half the bill.

#finops#agents#projects

introducing gh-aw-fleet →

Why I built a declarative fleet manager for GitHub Agentic Workflows: one fleet.json to deploy, sync, and track drift across every repo, plus what's shipped through v0.2.0.

#projects#agents#ai#finops

where the on switch is →

On the gap between people deep in AI and people still looking for the on switch, and why I'm going to keep writing for both.

#ai#meta

where did the knife come from →

A third-grade peanut butter sandwich assignment, and why it explains what people get wrong about prompt engineering.

#ai#agents

hello, world →

Why I'm starting a blog, and what to expect.

#meta