Welcome: run a study, inspect the evidence, contribute

[alohays]

Welcome to Allagma. Start with the offline first study: no model account or GPU is needed. Inspect the attempts, claims and deterministic audit, then share what you could or could not reproduce, including your environment and the limits of the evidence.

Looking for a first contribution? The pinned contributor overview links the milestone and three available, unclaimed tasks:

• Explain duplicate-key validation and recovery.
• Describe the EMA figure accessibly.
• Qualify the offline tutorial on Linux/Python 3.13.

Comment on a task with your intended scope and environment so the maintainer can confirm the claim. The three deeper implementation PRs are drafts awaiting owner review; their proposed features are not yet available on main.

Use Q&A for usage questions, Show and tell for study packages with source/evidence links and limitations, and Ideas for bounded proposals. Reproducible defects and accepted tasks belong in issues. Read the support guidance and code of conduct; report vulnerabilities through the private security channel. Keep credentials and private research data out of posts. No response-time guarantee is promised.

This is a maintainer welcome prepared with agent assistance, not a claim of external adoption or independent scientific qualification.