shipping/JKT — 03:41 GMT+7/48.3 °C CPU/v 07.0 · build 2026.04.20
lat −6.200 · lon 106.816/ntwk · online/open for Q3 '26
[ essay no. 05 ]craftapr 06, 202611 min3,104 wordsrevision 2live
don't trust the agent. audit the plan.
why i always show the agent's plan before it runs — and a bias toward visible reasoning.
d
devrangga hazza mahiswaracreative engineer · jogja, id
full essay coming soon.
[ if this moved you ]
keep reading.
three essays in the same key · pick oneagentsapr 14 · 26 · 9 min
the case for settling attention onchain.
micro-royalties priced in milliseconds. why streaming and onchain primitives converge twice before 3g is dead.
read essay →agentsdec 18 · 25 · 14 min
a year of small model surgery.
every swap i made inside a production agent pipeline in 2025 — what moved the needle, what didn't, and the three prompts i'll keep paying for in 2026.
read essay →toolsdec 22 · 25 · 4 min
trace · a cli agent that plans.
announcing the open alpha. three demos, one odd opinion about why cli agents still matter in a chat-first world.
read essay →