Field Notes
from Production
Random thoughts, occasional insights, and lessons from breaking things where the stakes were real. 47 pieces; probably a few should have been tweets.
Featured
Start product recommendations with rules, not AI.
Start product recommendations with your best sellers and clear rules. Separate customer and product eligibility from ranking before adding an AI model.
See which page ranks for each query with the Search Console API.
Use the Search Console API to retrieve page and query together and check which URLs appear for each term. The returned rows can reveal overlap between your pages.
Pricing rules when a guest becomes a known customer
Decide what happens to a selected offer after login reveals the customer's eligibility. Explain any price change and preserve the accepted offer through checkout and billing.
A checkout countdown has to earn its delivery promise.
A checkout countdown can create urgency or support a real delivery promise. That promise needs clear rules for what changes at zero, which date to show, and how it survives payment.
Codex vs Claude Code: two agents on one codebase
I run Claude Code and Codex daily, sometimes on the same repository. My session logs show the context cost, while the working tree shows the coordination cost.
Customer accounts: separating users, owners and permissions
A person may manage a service they don't own or place an order for a company. Keep the user, customer, permissions and order context separate.
What using Claude Code actually looks like
Using Claude Code is a loop: ask, change, check, decide. Here is what a normal session looks like, six commands worth learning first, and how to recover when it goes wrong.
Capturing Outlook email and calendar with Power Automate
Microsoft Graph and IMAP were off the table, so I used Power Automate, OneDrive sync, and one 15-minute scheduled task to pull email and calendar into a local inbox.
Splitting an AI pipeline between Python scripts and the model
Python handles parsing and deduplication in my inbox pipeline. The model gets the prepared text and extracts summaries, decisions, and action items.
Claude Code skills: building a workflow that runs itself
The control layer of my work system is markdown and small scripts. How a workflow combines Claude Code skills, rules, model routing, and parallel sub-agents.
The Claude Code plugins I built and run
I put the Claude Code tools I reuse across projects into one public toolshed: generators, workflow commands, anti-slop checks, search visibility, rule packs, and prompts.
My AI was smart. It just couldn't see my job.
My work context was scattered across email, meetings, documents, and notes. So I wired it into a private Obsidian vault Claude Code can read.