9 sessions on clang with a coding agent: the agent failed at least once in 9, and 0 finished without a human stepping in.
Takeaway: Next time I'd ask upfront how this OS's other interactive JS apps handle input, since the busy-poll getKeyNonBlock() loop turned out to be the real cause of the exit crash, not the VGA I/O port theory the agent chased first.
Takeaway: Next time I'd tell it upfront that Stage 3's 'parser/bytecode boundary' can't be a real physical separation in a 63K-line monolith like quickjs-ng, so it skips straight to building the JS_WriteObject/JS_ReadObject wrapper instead of discovering that mid-session.
Takeaway: Next time I'd spell out exact screen resolutions and the grid/sidebar split ratio in the very first prompt instead of iterating three times on layout — that back-and-forth cost more turns than any of the actual milestone work.
Takeaway: I'd start the first prompt by explicitly naming the root directory and the include/src split up front, since correcting those after the fact cost two early interruptions.
Takeaway: Next time I'd specify upfront which CLI flags must stay wired through even if the C API doesn't support them yet, since dropping --recover during the refactor silently broke a fixture test and cost a round of debugging.
Takeaway: Next time I'd ask upfront for an audit of what quickjs-ng already exposes (JSMallocFunctions, JS_WriteObject/JS_ReadObject) before specifying a stage's API, since the agent ended up discovering mid-stage that the 'ideal' architecture didn't match the 63K-line monolith's reality.
Takeaway: Next time I'd ask upfront for the stack-usage rules (no post-init malloc, bounded everything) to be double-checked against test harness patterns, since two of the three real bugs were stack overflows from large structs being copied or placed on the call stack instead of static storage.
Takeaway: I'd open with the exact directory layout (include/ for headers, src/ for cpp, root at d:\Yggdrasil) since that's the first thing I had to correct.
Takeaway: Next time I'd specify upfront that MB1 must stay fully working as a real fallback, not just compile — that black-screen/no-console debugging loop cost the most time.