40 posts
Run a monthly drill with a deliberate violation. Verify which controls reject it, report failures, and confirm cleanup. The drill attempts real merges, including an administrator override; use a disposable test repository and inspect the…
Run the task queue on a schedule. Separate urgent failures, requests for your decision, and routine logs. Detect missed runs, and understand why a watch on the same machine cannot detect that machine's complete failure.
Add a separate reviewer, require automated checks before merging, and verify the deployed result before marking a task complete. Configure a runner, a GitHub workflow, and command restrictions, with explicit limits on what each can verify.
Give each task acceptance criteria, dependencies, and a status. Define how the agent selects the next eligible task and when it must stop. Includes a queue template and a five-item exercise.
Write down the actions agents may take, the actions requiring your approval, and the evidence they must provide before reporting completion. Includes a permission template and exercises that check whether an agent follows it.
An eight-minute vibe-coding challenge at a Mouse AI League mentor mixer, where I volunteer as a mentor and judge: an AI agent built a French flashcard app in eight minutes, and the speed of the machine is what makes product judgment the…
I cap my setup at roughly ten coding agents across several subscriptions, with three or four per vendor. A seven-agent wave exposed failures in network access, checkouts, test ports, and generated files. The counts and the preflight chec…
Ballast declines 24 of its 89 golden questions on purpose and answers the other 65 with citations. As of the August 16 ledger the answer-or-decline call was right on all 89, and one of the 65 answers was unfaithful to its sources. Where…
Thirty days of running an agent fleet from chat: 8.7 billion tokens, 70,227 calls, 2,861 sessions, a 98 percent cache hit rate. At each vendor's list price that traffic bills $7,566. Three subscriptions covered it for $599. Where the tok…
Document which repository owns each part of the project. Assign each agent a specific scope and a separate checkout to prevent overlapping edits. Includes layout instructions and a repository-splitting exercise.
Turn a project rule into a script that detects violations. Run it before each push and test it with a deliberately invalid change. Learn what a local Git hook checks and how it can be bypassed.
Write a decision record documenting what you chose, which alternatives you rejected, and when to reconsider. Instruct agents to read relevant decisions before proposing changes, then test whether they retrieve them.
The third post in the SDD numbers series is not about velocity. Since June the platform's effort moved into four quiet systems: an externalized memory, a statistical screen that refuses to flatter me, a point-in-time data floor, and a pr…
The full path to an agent platform that edits its own instructions is already published, scattered across seven articles that never mention each other. I run that architecture as one person: supervisor and worker tiers, verification guar…
A Windows Task Scheduler default fed my Claude Code memory pipeline its own transcripts for nine days: 21,264 junk sessions, 38 GB of exhaust, and an 18,428-session backlog that was really 1,924. The durable fix: postconditions that outr…
I split a day of development across two AI subscriptions: judgment on the Anthropic meter, typing on Moonshot's. Five merged PRs later, here is where the tokens went, how K3's pricing actually compares, and how I'm tuning the mix.
Record what a feature must do, what is outside its scope, and how you will verify it. Create a specs directory and template, then give the agent an approved spec to implement.
I built a method as a public repo and the product that runs it as two private ones, and none of them treated the others as a source of truth. The domain enum lived in four places. A persona drifted between its lens file and its API contr…
Three months ago I laid out spec-driven development and the folder architecture that makes it work. Most of the playbook held up in daily production use. Three ideas that essay never mentioned turned out to matter more than anything in i…
Spec-driven development, pointed at a life. Why my principles and goals live as markdown an AI agent reasons over, and the one rule that makes handing an agent your life safe: it can challenge the record, but it never writes it.
I built a quant research platform, then built an agent to operate it: a scheduled Claude session that reads the boards, keeps a pre-registered track record, and texts me three times a day without ever saying buy.
I built a self-healing RAG pipeline, a guardrails gateway, and an eval gate as one system, then threw a 44-question battery at it. Zero hallucinations, because the behavior that matters most is refusal. Here is how trust got built into t…
A year of building and operating a small fleet of finance and content products almost entirely through an AI coding agent. What worked, what was hard, the honest failures (including a flagship signal that measured nothing and an edge tha…
Prompt caching looks like a flag you flip for a cheaper bill. It is really the reuse of a stored prompt prefix, governed by three rules, and applying it across four parts of my own system showed where it pays, where it quietly does nothi…
Every system that fuses signals into one consequential number has a fault line: the data you trust enough to composite into a grade versus the data you only trust enough to watch. How I drew that boundary in my personal finance engine, a…
Four days after I said goodbye to Opus, an export-control directive pulled Fable 5 offline and the fallback became the workhorse again. What I shipped in the window, what it cost, and the model-tiering plan for when Fable comes back.
I handed a backlog to Claude Fable, told it once it could merge, and let it run. It shipped seventeen items across five repos. The line that mattered was not in the work it finished. It was in the work it refused to touch.
Anthropic shipped Claude Fable 5 and Mythos 5: same model, two names, one safeguard layer apart. What the new frontier model means for running agents in production.
Dumping the whole corpus into an AI agent makes it worse, not better. The fix is architectural: each task loads a curated slice, not everything you have. Here is the method, and the same move at three different layers: specs, sensor data…
A small ARM box that started as a local LLM experiment and ended up a self-governing node: private retrieval, a resident agent under a written constitution, a code-enforced safety fence, and a nightly job where it audits itself and files…
Anthropic published a guide on building a session-level orchestration mode. I built it two ways, on the CLI and on the API, and then hit the part the guide does not cover: an orchestrator that fans out is useless without a backlog of rea…
How I replaced manual CSV exports with a live Garmin data feed for my AI marathon coach: a scheduled unofficial-API poller, resilient session handling, and the design calls that keep training and recovery data fresh and trustworthy.
Two posts ago I bet that keeping my portfolio reviewer's engine deterministic and auditable was worth it. This is where that bet paid off: because the engine is replayable, I could run a simulated market crash through the real production…
A personal portfolio reviewer where the scoring is deterministic and the AI only narrates. The architecture that held up after I had to rewrite the model it was built on, and why that boundary is the whole point.
Two weeks after I shipped a post about a scoring engine I'd built, I rewrote the spec it was based on. Here's what I learned, and why I had an AI agent do the literature review.
Spec-driven development reads like a methodology for controlling AI agents. It isn't. It's a methodology for managing context across stateless sessions. The spec is the persistent memory.
Two production sites, a blog, and two personal AI projects, shipped this week from a phone. The chain is voice dictation into Perplexity Computer, a spec, then Claude Code on the web. The interaction model is the story.
How I built a personal AI coaching system for marathon training, layering deterministic guardrails over an LLM narrative engine, ingesting Garmin FIT files, and designing for my own injury history.
Why spec-driven development and structured folder architecture are the missing infrastructure for AI-assisted engineering: methodology, common mistakes, and where to start.
My review of Doug Kerwin's Enterprise Vibe Coding Playbook: its advice on prompting, design, and review, who I'd recommend it to, and the spec workflow I'd add.