Vol. 1 · No. 14 · Monday, November 9, 2026
Build logs from shipping MacGet and AI systems in public — the engineering, the trade-offs, and what broke along the way.
Seven steps to turn a repo-local Claude Code command into a real plugin: plugin.json, a SKILL.md, claude plugin validate, and a marketplace, using my /build-log command as the example.
Nov 5, 2026 · 5 min read
Earlier editions — 13
Signal Boy: a zero-dependency WiFi survey tool you run with npx
I wanted to find the dead zone in my flat without installing a heatmap app or signing up for anything. So I built Signal Boy — one npx command, a live dBm readout on a Game Boy LCD, and no dependencies at all. Here's what that constraint cost me.
Aug 7, 2026 6 min
Shipping an MCP server is the easy half. The work that decides whether an agent can actually drive your product is payload design — closed vocabularies, schema-constrained returns, and tool descriptions written for a reader who will not ask a follow-up question.
Aug 3, 2026 6 min
MacGet 1.3.0 adds BitTorrent and OPDS book catalogs to the macOS download manager — and rips out the adaptive connection ramp I shipped as a headline feature in 1.1.0, because the heuristic read a saturated link as a failed experiment.
Jul 30, 2026 7 min
I wrote a Node script to generate whiteboard-sketch covers for every post on this blog, and banned the model from drawing readable text. Nano Banana Pro ignored the ban, spelled everything correctly, and proved the constraint was a leftover assumption from 2023.
Jul 26, 2026 5 min
Opus 5 landed at half Fable 5's price and claims to approach it on most tasks. For a multi-agent research desk the real question isn't which model wins — it's which node gets which, and the two API changes that will silently break your existing graph.
Jul 26, 2026 6 min
How I added a full build-log blog to my portfolio — likes, email subscribers, and new-post notifications — without paying for anything. The stack, the code, and the one Vercel Hobby limit that broke my cron.
Jul 22, 2026 6 min
The system behind an AI that reads filings and answers like an analyst — specialized agents on LangGraph over a Vertex multimodal RAG data plane. What the architecture looks like, and the four things that broke on the way there.
Jul 16, 2026 8 min
Prompt injection isn't one problem, so it doesn't get one fix. How I layered deterministic hard gates and an LLM-as-judge soft gate across a LangGraph agent — and why the refusal itself turned out to be the leak.
Jul 15, 2026 7 min
LangGraph runs the control plane; Vertex AI's managed multimodal RAG Engine runs the data plane. How I got retrieval quality and groundedness where I needed it — because in financial filings, the number lives in the chart, not the paragraph.
Jul 12, 2026 7 min
The context-management habits that keep an AI coding agent sharp over a long session — when to /clear, when to /compact, and how to write a CLAUDE.md the model actually reads.
Jul 10, 2026 7 min
The infrastructure lessons behind running LLM features in production — streaming, caching, cost routing, and treating the model as an unreliable network call.
Jul 10, 2026 5 min
The structure I reach for when building production AI agents — explicit state, small nodes, tool boundaries, and knowing when NOT to let the model decide.
Jul 8, 2026 6 min
A build log on how MacGet downloads files fast on macOS — chunked HTTP parallelism, an actor-based engine in Swift, browser-capture over Native Messaging, and bundled yt-dlp for video.
Jun 27, 2026 7 min