We said MAYBE on August 13, 2026. Not settled — due August 13, 2027.
Read this with the caveat. We tested the engine that produced this verdict against 292 launches whose outcomes we already knew, and could not show it predicted which survived. Some of its data sources were also dead at the time of scoring. The verdict stays up, dated and unedited, because a record you can quietly revise is not a record — but it is worth less than it looked when it was written.
Demand for coding agents is loud and commercially proven, but the market is dominated by well-funded giants and building a competitive agent is expensive engineering. The only viable path is a narrow game-dev vertical wedge built on top of existing open-source agents/models — not a head-on coding-agent fight.
Multiple high-scoring HN discussions (1274, 1274, 1036 points) plus four established competitors charging money proves strong, current demand, and search direction is growing.
A real vertical wedge exists (game-engine specialization, offline model, itch.io/game-jam marketing untouched by incumbents), but the incumbents are funded giants and the core coding-agent engineering is heavy, blunting solo beatability despite the niche.
The scored axes lean SKIP on winnability, but the strongest case for building is that generalist incumbents genuinely ignore the game-dev vertical: no competitor markets to itch.io or offers Unity/Godot-tuned prototyping, and a solo dev with game-engine domain knowledge could own that niche as a thin IDE plugin layered on existing open models (OpenCode/Qwen) rather than building an agent from scratch.
What is worth more than this page. The register records what became of 6,266 real launches. No engine has to be right for that to be true.