We operate a fleet of repositories where agents write most of the code. The failure mode we fight most often is not wrong code. It is extra code: the speculative flag, the abstraction with one caller, the error handler for a state the system cannot reach. A human engineer gets
In July 2026 we promoted a decision into the fleet's canon: learning from skill execution happens invisibly, while the skill runs, with no scheduled review and no ceremony. A learning is either detected in the moment and passed through four quality gates, or it is not kept at
Every repository in our fleet runs on Loa, the operating layer of slash commands that carries a project from first idea to deployed artifact. Loa grew the way most command surfaces grow: one honest, specific name per phase of work. Planning got /plan-and-analyze, then /architect, then /sprint-plan. Implementation got /implement.
Our agents do not read their inputs first. A pre-execution layer reads them first, and that layer has authority to stop the run before a single tool call is issued. The specification is short enough to fit in one line of our operating doc: PII filtering (blocking), injection detection (blocking)
One of our operating records is four words long in substance: prompts are automatically enhanced before skill execution, silently, and logged to the trajectory. It reads like an implementation note. It is closer to a constitutional question, because it decides who knows what the fleet was actually asked to do.
The fleet is a small number of repositories where most of the writing is done by agents and most of the reading is not done by a human at all. That inverts the usual bottleneck. Producing a change is cheap. Deciding whether the change is allowed to land is the
There is a one-line rule in our Loa, the standing instruction layer every agent in the fleet loads before it touches a repo. It reads, in full: for source files, ALWAYS use the Write tool. The capitals are in the source. It was captured from our docs and promoted to
Every repo we run carries a conventions block near the top of its operating instructions. In the Loa framework that governs our fleet, that block is three lines long: Never skip phases. Each builds on previous. Never edit .claude/ directly. Security first. Three lines is not an accident of laziness.
Every repository in our fleet gets visited, many times a day, by an agent that has never seen it before. That is not an edge case for us. It is the normal condition of operation. A fleet run by agents is a fleet of cold starts, and the first thing
Some of the most consequential decisions in a codebase are the ones nobody remembers making. Where a file lives is one of them. Loa, the framework layer our repos share, kept its grimoire and its state files at fixed locations for a long time, and those locations quietly became load-bearing.
Our loa layer carries a promoted decision record. Stripped of its provenance block, it is three sentences long: Beads task tracking is the EXPECTED DEFAULT. Working without beads is abnormal. Health checks run at every workflow boundary. There is no benchmark in it, no adoption target, no migration timeline. It
People sometimes ask how a fleet of agent-run repos produces videos and social posts on a steady cadence without a content team. The honest answer is that there is no muse anywhere in the building. There is a pipeline, and the pipeline is made of small, inspectable mechanisms. This post