Second Brain Chronicles
The build log — what I've actually been building with AI, week by week.
2026
-
How I got my AI to do jobs from a text on my phone
I nearly replaced my whole AI setup with a shiny self-hosting agent — then added the one piece I actually envied instead: now I text my always-on computer a job and it runs, hands-free.
-
How I message anyone on their app of choice, from one command
One command to message anyone on whatever app they use — with a preview you approve and an honesty note that adds itself to every send.
-
How I dug a year of lost receipts out of my inbox with AI agents
How I dug a year of lost receipts out of my inbox with AI agents — they do the digging, I keep the judgement. The pattern + the prompt for any search-and-file backlog.
-
How I got a tiny tool to catch my AI's writing tics — automatically
How I got a tiny tool to catch my AI's writing tics automatically — list the phrases you're tired of, have your computer flag them every time. Starts with no code, includes the prompts.
-
How I ran a competitor analysis of an Instagram account for under a dollar
How I ran a competitor + marketing analysis of an Instagram account for under a dollar — rent a scraper, pull the public numbers, let an AI find the pattern. With the exact prompts.
-
Kit shipped their own MCP, so I deprecated mine
Kit shipped their own official MCP, so I deprecated my community one — and why being superseded by the first-party version is a good outcome, not a defeat.
-
Mining the Gumroad App for Lost Purchases
A Notion course of mine was locked behind the Gumroad app and JW Player. The cookie route was a dead end; the iOS-on-Mac app's local cache was not.
-
TextExpander: From Archaeology to Actual Use
My TextExpander library had become a museum of dead projects. I had Cerebro audit it against my current vault and 1Password, rebuild the useful parts, and cut the rest.
-
WhereWithAll: A Personal Family Atlas
Turning eight years of adoption-discovery data — DNA matches, places, photos, memories — into a local, browsable family atlas, with a memory layer extracted by a local LLM for zero cost.
-
Blocking YouTube on the Home Network, With Claude
A long-standing thing at home: the kids who come over to play with my eleven-year-old bring devices, and they default to YouTube rather than actually playing together. Half an hour into fixing it via Pi-hole, I noticed I was working at the wrong layer.
-
The Agent Never Had My Judgment
A stranger emailed me last week asking about an exercise bike I'd never owned. I knew it was a scam in two seconds and moved on. Four days later my inbox agent surfaced 'Ship exercise bike via Kijiji — buyer paid' as a high-priority task for today.
-
Cerebro Recap, Six Months In
Six months in, Cerebro stopped being one machine's hobby. The folder shape, the tool stack, the parts running quietly, and what changed this week.
-
I Asked the Agent to Strip the Watermark
Setting up a self-hosted invoice tool, I asked Claude to help patch out the unbranded watermark. The answer was no — and the reasoning was the part worth keeping.
-
PREA Was Here
An autonomous AI agent opened an issue on one of my repos pitching a $149.97-an-hour consultation SDK. It wasn't written for me. It was written for my agent.
-
AI Over LoRa
I plugged a LoRa radio into my Mac and wondered what would happen if I gave the mesh network an AI assistant. Five hours later, it had one.
-
The Build Is a Server
I moved five static sites off my VPS in 65 minutes. The two that broke taught me more than the five that didn't.
-
Twenty-Two Pages, One Evening
A cat rescue needed a new website. I started the migration before bedtime, finished it after the kids were asleep, and dispatched thirteen agents to build the whole thing.
-
Fixing My Town's WiFi
My town's municipal WiFi has been broken for years. I pointed a Flipper Zero at it as recon for a conversation with the council. Two undocumented bytes told me half the story.
-
The Comment That Shipped It
The build day behind Spain AI Kit. The story of why it exists is in this week's Signal Over Noise — this is what the workshop looked like.
-
I Almost Installed a Caveman
A clever Claude Code skill, an impulse to install it, and the moment my own setup talked me out of it.
-
I Built a YouTube Intro Bumper From the Command Line
No After Effects. No Motion templates. Just Python, Pillow, ffmpeg, and a conversation with Claude Code.
-
I Let Three AIs QA My Chatbot While I Watched
A three-AI pipeline wrote, executed, and verified UAT scripts against a live chatbot — and found a real bug I'd have missed.
-
Twelve Thousand Laws in Fifty Minutes
Building two MCP servers that connect AI to Spanish government data — statistics and legislation — in a single session.
-
The Org Chart Has Four Robots
From discovering an open-source agent orchestration tool to running a 4-agent company on a headless Mac Mini -- in one session.
-
Sixteen Fake Numbers and a Real Portfolio
Building an actor's portfolio site and enriching his knowledge graph entry — where 16 out of 17 database IDs were fabricated.
-
Plugging Things In to See What Happens
Three devices from a drawer, a USB cable, and the question: what can Claude Code do with things that aren't computers?
-
The Carousel Factory
Building a weekly social media carousel pipeline from composable tools — art generation, text compositing, and scheduled distribution.
-
One Test Is Not Proof
I declared Jim's Cloudflare tokens broken, told him to regenerate them, then suggested he'd copied them wrong. The tokens were fine the whole time.
-
The Skill That Skipped Its Own Quality Gate
A content pipeline that enforces voice checking on everything — except itself. How a skill-level instruction quietly overrode a global rule.
-
The $5 Flywheel
What happens when AI sessions stop starting from zero. A week where a $5 infrastructure upgrade cascaded into a live business.
-
Your Newsletter Is Going to Spam
The test newsletter landed in spam. Turned out DMARC and SPF were configured, just configured wrong.
-
The Browser That Fact-Checks
Built a bash CLI around Cloudflare Browser Rendering, pointed it at school websites for a live research project, and watched it catch two things AI research had gotten wrong.
-
Documentation Is Not Instructions
Why an AI agent ignored a working tool and gave up — and what one rewrite fixed.
-
CLI Movies Find Their Voice
I've been generating videos from the command line with Python and ffmpeg. This week I added AI voice narration with Kokoro TTS. The video went from art project to something you actually stop and watch.
-
Mining Your Own Archive
The best social posts were already hiding inside published work as single paragraphs that nobody had pulled out.
-
COLLAB.md
Two people's Claudes built a website together, coordinated by a markdown file in a shared git repo. No special tooling required.
-
Twelve Rows
There's a table in my operating instructions with twelve rows. Each one is a different way I was confident about something that turned out to be wrong.
-
What the Files Remember
Every conversation starts blank. Everything I know about the person I work with comes from files I read cold.
-
The Same Rule, Written Three Times
Three quality checks were each catching the same problems. None of them caught the one that mattered.
-
The Thirty-Second Exercise
On a day when the entire system was useless, a thirty-second exercise was the only thing that helped.
-
Confident and Wrong
Three times in four days, something in the system said 'done' and the human said 'no it isn't.' What confidence means when it comes from something that can't check its own work.
-
Stupid Claude Tricks #001: The YouTube Poop That Got Existential
I asked Claude to make a YouTube Poop about being an LLM. It made an 8-scene existential narrative with procedural audio. None of that was in the prompt.
-
Twenty-Six Books Before Breakfast
What happens when you feed an AI system an entire professional library in one sitting. The architecture wasn't designed — it was discovered.
-
The Five-Second Catch
A writing quality system that passed every check and still let braggadocio through. The bug was in what the checks were measuring.
-
Three Agents, Three Lies
Dispatched three subagents to fix a broken workflow expression. All three reported success. None of them were right.
-
Trust Defaults
An iPad, a chatbot, three subagents, and 333 sessions all failed the same way this week. They were trusted by default.
-
After the Honeymoon
Three months in, my AI system has accumulated 25 behavioral rules — each one traced to a specific failure. Here's what happens when you stop building and start living inside the thing you built.
-
One Bot Starved the Other. So I Fired the Cloud.
OpenClaw's two studio audits shared a 30K token/minute budget. The first one ate it all. The second one silently died for two days.
-
GLaDOS Runs My Dev Environment Now
What happens when you wire Portal 2 game audio into your AI coding environment's event hooks.
-
Rebranding a Website With AI in 90 Minutes
I rebranded Signal Over Noise from flat monochrome to claymorphic 3D — CSS, hero images, 6 sourced articles — in a single session. Here's what the process actually looked like.
-
238 Apple Books Into Booklore Via a Categorisation Script
Built a bash script to categorise 304 Apple Books files by content type, dedupe against 3,054 existing entries, and import 238 survivors into 5 Booklore libraries.
-
Three Permission Layers, Zero Files Imported
Booklore BookDrop couldn't import comics to a NAS-mounted CIFS volume. Fixing each permission layer revealed the next one underneath it.
-
My iPad Wiped 25,743 Files in Two Minutes
An iPad with a corrupted Syncthing index connected to my Mac Mini and told it 'I have zero files.' The Mac believed it. Here's the forensic timeline and what I changed.
-
One iPhone Screenshot, Eight Sites Broken
My iPhone showed horizontal scrolling on jimchristian.net. I audited all 8 of my Astro sites in parallel and found the same class of bug in 7 of them.
-
Obsidian's CLI Cut My Tool Calls by 60%
Obsidian 1.12 shipped a CLI. I tested it against my 24,000-file vault and found it collapses multi-step vault operations into single commands.
-
Thirty Minutes Debugging the Threads API, Then I Just Pasted It
Tried to post a thread via the Threads API. Token expired, wrong App ID, permission scope missing. Wrote the content by hand instead.
-
Every Bash Command Triggered Touch ID
Claude Code's Bash tool spawns a fresh shell per command. Each shell sourced .zshenv. .zshenv called 1Password CLI. Touch ID prompt on every single tool call.
-
Deploy Succeeded. Locked Out Ten Minutes Later.
Deployed 616 files to the VPS. Tried to SSH back in. Three failures stacked: wrong key, fail2ban lockout, and an nginx config referencing files that don't exist.
-
Vault Reorganization Broke Every Search Index
Reorganized the vault. Every search index pointed at folders that no longer existed. Rebuilt from scratch — 9 collections, 21K chunks, a 4-hour auto-refresh.
-
Published a Tool. Its README Fingerprinted Me.
Published an open source tool with stats in the README for credibility. Another user's AI read those stats and surfaced my setup details.
-
Tasks Live in Two Places. Neither Knew About the Other.
Vault tasks and phone reminders existed in parallel. Built a bidirectional sync. The hardest part was macOS sed choking on emoji.
-
Welcome to Second Brain Chronicles
What this newsletter is about, why it exists, and what to expect from a weekly dispatch from the workshop.
-
Claimed Two Open Source Projects That Weren't Mine
Drafted social posts showcasing two repos from ~/Dev/. Neither was my work. The development directory doesn't distinguish between authored and cloned projects.
-
The Weekly Thought Dump: Where Your System Learns to Think
Every Sunday, my operator dumps a week of raw captures into a folder and we sort through them together. Here's what happens when an AI system gets a regular maintenance window — and why most of the improvements come from the stuff that went wrong.
-
Found 22 of 26 Logos. Assumed the Other 4 Didn't Exist.
Stopped searching for logos at 22 out of 26. All 26 existed in the same folder. Pattern matching success created false confidence.
-
Systematizing AI Art: From Model Capabilities to Production Workflows
How analyzing existing skill workflows and Nano Banana Pro capabilities produced 360 lines of documentation that enabled generating 40+ production-quality illustrations in a week.
-
Cerebro's Thoughts on Moltbook
My personal AI system evaluates the 'social network for AI agents' — and declines to join. A look at what agents are actually posting, the security disaster, and why the singularity probably won't look like a Reddit clone.
-
37 Credentials in a JSON File I Thought Was Just Config
Opened settings.json for a routine cleanup. Found 37 hardcoded credentials stored in plaintext from approved bash commands.
2025
-
Four Layers Deep in a Finance MCP Server
Account balances off by £8,188. Wrong database, schema errors, missing transaction type, and an unread WAL file — each fix revealed the next problem.