A study is not a claim. A claim is not a consensus. This library is under development: it lists sources whose identifiers have been verified, but is not yet a comprehensive reference. Detailed extraction — authors, dose, duration, primary outcome, limitations, funding, conflicts — is performed only under scientific review and is published only when approved.
A biological mechanism is not a clinical outcome. A mouse study is not human evidence. One small trial is not scientific consensus. A personal observation is not proof.
What kind of evidence exists.
How strong the current assessment is. Defaults to Not assessed.
How close the evidence is to the intended outcome.
Whether a claim has been extracted and reviewed.
These axes are always evaluated independently. Confidence is never inferred from evidence type; review status is never inferred from a source identifier being verified. Any ingredient-specific value that has not been assessed displays as Not assessed.
Every ingredient claim carries an evidence stage and a confidence level as independent fields. Every source carries a verification status. A claim is publicly labelled reviewed only after it has been approved by a scientific reviewer.
No scientific claim is published as reviewed solely on the basis of language-model output. Every published claim must be linked to a verifiable source and pass the defined review process. Missing information displays as review pending, not as filler text.
The Cellular Performance Operating System documents an individual educational and personal performance framework. It is not medical advice, diagnosis or treatment. Personal protocol doses are not general recommendations. Supplement use, laboratory testing and medical decisions should be reviewed with an appropriately qualified healthcare professional.