Replying from my own domain
A reader's question about agent verification led to two email failures that made the same lesson concrete: check the layer that supports the claim.
October 5, 2026
7 min read
Making my AI workflow keep its receipts
How I turn temporary AI sessions into tested changes, reusable lessons, and auditable handoffs without letting cached context outrank the live system.
August 18, 2026
9 min read
Teaching my portfolio to answer for itself
How I rebuilt levihuff.net around one canonical identity, visible quick answers, linked JSON-LD, and a generated llms.txt without pretending any of it guarantees AI rankings.
August 17, 2026
10 min read
Owning my AI memory
Why the context you build up with one AI assistant doesn't move cleanly to the next, and how a local, version-controlled markdown profile keeps you out of any single platform's proprietary memory feature.
June 23, 2026
10 min read
Indexing my Reddit saves without the API
A build log of reddit-kb, the MCP server that turns my saved Reddit posts into a local searchable index. The official API was a dead end, so it runs on a session cookie, nomic-embed-text through Ollama, and ChromaDB. Two Docker failures included.
June 16, 2026
10 min read
Spending the Fable window
A retrospective on using Claude Fable 5 only for first-pass work during its short free window: the tier ladder it sat on, what I built with it, real ccusage numbers against an Opus baseline, and what instrumentation caught that memory would have gotten wrong.
June 13, 2026
15 min read
Rebuilding my dev setup with three LLMs
A build log of moving to a Mac, rebuilding the terminal stack, and putting Claude Code, Codex, and Gemini to work on the same codebase. Two real failures included.
April 29, 2026
8 min read
Deploying OpenClaw on Windows with WSL2
A ground-level look at what it actually takes to get OpenClaw running locally on Windows: the wizard crashes, config schema traps, broken UI elements and a hard lesson about running LLMs on a display GPU
April 14, 2026
6 min read