Newest first. Tagged by what it concerns. Each entry links to the post, and the post is the durable record.
Four months of closed alpha with ten design partners, and a pipeline renamed by the people using it. Beta opens applications beyond the alpha cohort. Every claim still keeps its receipt, every persona is still inspectable, and nothing runs until you approve the question.
Readcreativehuman.ai is rebuilt: services on top, the platforms underneath, the record in one place. Before the rewrite shipped we put our own go-to-market question through Glasshouse. What came back is in this note, confidence caveat and dead citations included.
ReadOne free-text request. One Q&A round. End-to-end autonomous execution: plan, decompose, parallel execute, verify, hand back a PR. Plus the always-tested rule: 100/100/100/100 coverage on every changed line.
ReadEighteen months of internal use, opened today to ten design partners. Decision rehearsal against a simulated public: every claim footnoted, every persona inspectable, every run pinned to a canonical SHA.
ReadThe orchestration layer we run on top of Claude Code, Cursor and OpenCode, released under the MIT licence. 48 agents, 35 commands, one command from idea to tested PR. Runs on your machine; your code never leaves it.
ReadEight-week engagement, closed today. Custom momentum models, self-explaining analytics, continuous-evaluation loop, built for the brokerage-and-trading arm of Riyadh-based Zahran Holding.
ReadWant it built? We build fast and well on Brew Chai, and we teach your team to do the same.
Bring us a scenario