Built-in strategies, Strategy Studio, the bounded Research Lab, compare, backtest, paper tracking, scoreboards, guide/docs, and shared result links are the public-facing workflow.
The public product stops at Strategy Studio, bounded Research Lab, and result sharing.
Custom code remains an internal admin-only lab capability. Alphrex is not building a near-term raw-code community layer, and any public sharing that remains belongs on saved results and research outcomes rather than code distribution.
Settled Product Policy
This is no longer an open roadmap question. Public authoring is intentionally bounded, and raw-code community publishing is outside the current product claim set.
Raw Python, free-form custom strategy code, and internal snapshot registries stay admin-only until a real execution-isolation model exists. They are lab tooling, not public product pillars.
Sharing in Alphrex means publishing results, diagnostics, and takeaways through saved result links. It does not mean distributing raw strategy code to a public community layer.
Why The Boundary Exists
Bounded templates and bounded research tasks can be governed. Arbitrary code execution is a different safety problem and should not be marketed as half-open.
Strategy Studio and Research Lab constrain inputs, runtime behavior, and output shape. That makes them compatible with a public workflow and with the product’s actual trust ceiling.
Custom Python can still access failure modes that bounded templates avoid. Until execution is isolated properly, that surface remains a private admin lab rather than a public promise.
If public code execution ever returns, it should come back through real isolation or a materially narrower runtime. Until then, result sharing is the only public sharing story.