Every method we run, the limit it is judged against, and — the part usually left out — the question it does not answer. A test you misunderstand is worth about as much as a test nobody ran.
Not everything sold alongside peptides is a peptide. Small molecules take a different required panel, and applying peptide logic to them — as we have seen done on certificates in circulation — produces documents that are wrong in ways nobody catches.
| Test | Method | Specification |
|---|---|---|
| Identity | LC-MS against certified reference standard | Within ± 1.0 Da of theoretical |
| Purity | RP-HPLC-UV at compound-appropriate wavelength | ≥ 95.0 % area |
| Assay / content | HPLC against certified reference standard | ≥ 90 % of stated content |
| Water content | Karl Fischer titration | ≤ 5.0 % |
| Residual solvents | Headspace GC-MS | ICH Q3C / USP <467> |
This is the part of a testing programme that is almost never disclosed, and it determines what the result is worth. The four methods are not equivalent and we label which one applies every time.
| Method | What happens | Strength |
|---|---|---|
| Witnessed draw | Our officer draws units from the sealed lot at the filling site | Strongest — the supplier cannot select the unit |
| Random pull | Units selected at random from the sealed lot under a documented sampling plan | Strong |
| Retail purchase | We buy the product anonymously through the ordinary sales channel | Strong — the seller does not know it is being tested |
| Client-submitted | The client selects and sends the unit | Weakest — certifies that vial only, and cannot speak for the lot |
If we test a lot, every result we produce for it appears on this site — pass or fail. A supplier cannot commission a test and bury the outcome. This rule is the only reason anyone has cause to believe the passes.
A record missing any required test renders as INCOMPLETE and names what is missing. There is no configuration that turns a partial panel into a clean certificate.
Every chromatographic result carries its wavelength, column, gradient and retention time. A figure nobody can reproduce is not evidence.
No lot number, no record. A certificate that floats free of a batch will be used to sell everything the seller ever ships.
Issue date, age in days and retest date appear on every record. Results more than a year old are flagged as ageing rather than quietly reused.
Records live on attest214.com. A supplier cannot edit one, and cannot remove a failure from one. An image on a shop's website is not a record.