Gemini 4 Neon, Measured: My Test Broke Before Any Model Did
Gemini 4 Neon against Opus 5.5 and GPT-6.2 Astra on my own code: 155 runs, five of eight tasks dropped out. What holds: cost and speed, not a ranking.
5 posts on this topic.
Gemini 4 Neon against Opus 5.5 and GPT-6.2 Astra on my own code: 155 runs, five of eight tasks dropped out. What holds: cost and speed, not a ranking.
One lead session coordinates my parallel coding sessions with leases, constraints and a merge order. Open questions reach me in one queue. Now on npm, MIT.
Felix Rieseberg's depth prompt has been in my planning skill since 19 October. 43 past surprises: old flow 0 hits, the panel 11. Plus what it cannot do.
session-orchestrator counted placeholders as sessions. A test that never started could still pass a check. How the tool learned to check its own numbers.
Six screenshots sent on the go become four tasks. How Sven, Navigator and session-orchestrator work, and what I still decide.