Field Notes
agentic-sdlc-patterns
Two AI reviewers, independently, with real build and openssl access, found a bypassable revocation path and a spoofable certificate identity field in device-CA code. Both were confirmed with a live exploit, before and after the fix.
28 Aug 2026
agentic-sdlc-patterns
Five doubt cycles on one small registry guard, and the finding rate never reached zero. What actually caught the headline bugs was mutation testing and fresh-context review, self-inspection caught none of them.
27 Aug 2026
agentic-sdlc-patterns
A live-model coordination comparison looked decisive: peer-delegation succeeding, central-orchestrator starving the colony. An independent review asked whether it was actually a fair fight. It wasn't.
26 Aug 2026
agentic-sdlc-patterns
Adversarial review, applied at intent-declaration time instead of code-review time, caught real defects before a single commit existed to review, proving the doctrine is a decision-making technique, not just a code technique.
25 Aug 2026
agentic-sdlc-patterns
Three adversarial review rounds on the same new codebase, and the diagnostic that separated a durable fix from a fragile one: did it change a rule, or just a line?
24 Aug 2026
agentic-sdlc-patterns
Four documented review cycles, one recurring shape: a fix, left unreviewed by a second perspective, quietly introduces the next bug. The sentence the factory arrived at after proving it the hard way, four times.
23 Aug 2026
Five doors, not nine categories
The honest, chronological story of building the factory.
Reusable ways to deliver work with agents, including agent roles and model choice.
The shape of the system: context infrastructure and architecture at working size.
Controls that keep the work safe without killing momentum, and the logic for promoting or retiring ideas.
How one human actually runs the whole thing day to day.
Arcs
7 posts
One week, one question: can a 24GB laptop running only local models safely fix real code? The honest answer went from "it cannot produce a single valid patch" to "two local labs autonomously fix each other's code" in seven days, not by making the models smarter, but by repeatedly moving trust from model instruction-following to mechanical verification.
9 posts
One question, asked after a project spent 25 days producing documentation and no working code: why does a factory that can build individual tools still struggle to coordinate them into one real product? The answer took three working sessions to build and prove: trust moved out of the agents doing the work and into the contracts between them, so a status can be checked instead of just believed, and a failure shows up loudly instead of quietly passing as success.
6 posts
The factory's adversarial-review practice didn't start as doctrine, it became doctrine because the same failure kept recurring: a fix, itself unreviewed by a second perspective, quietly introduced the next bug. This arc is that doctrine's origin story plus four demonstrations that it holds outside its original context, closing on the sharpest evidence available: two real, live-exploit-confirmed bugs a second reviewer caught that the first pass missed.
Live ยท Factory Pulse
Fetched from GitHub on this page load, not baked in at the last deploy. Scoped to the specific repos that have graduated to public artefacts -- most of the actual factory stays private.