SYS · ONLINE│BATCH 2026·04·20│27 SAMPLES · 12 SUPPLIERS · 51 PEPTIDES│PASS · 63.0%│2026-04-20│v0.2.1
§ Standards
What we expect from vendors.
These are the substantive standards the Open Assay editorial team applies when reviewing a supplier. They sit alongside the eligibility criteria and the composite ranking formula. Publishing them openly is how a vendor knows what they are being judged against before we ever ship them a sample.
These are standards, not contracts. A vendor who hits every “pass” bullet below will earn a strong composite score when we test them. A vendor who matches any “fail” pattern will not. Editorial judgement is applied at the margins, in writing, on the vendor's supplier page.
§ 01
Five standards
OBJECTIVE RUBRIC
Measurable quality
Every sample we test must meet objective standards for identity, purity, and quantity against what is advertised, plus sterility and contamination criteria. We publish the underlying numbers, not just pass/fail.
What a pass looks like
- Identity confirmed by mass spectrometry against expected MW (±0.5 Da for [M+H]⁺).
- Purity ≥98% at 214 nm by reverse-phase HPLC.
- Quantity within ±10% of advertised mass.
- Endotoxin reported in EU/mg, even if no pass/fail threshold is applied at MVP.
What a fail looks like
- Wrong molecule at expected MW (identity fail).
- Purity under 98% or with major unidentified peaks.
- Quantity more than 10% below advertised; under-filled vials.
Batch consistency
Quality on one vial from one batch is weaker evidence than consistency across multiple vials and multiple batches. Vendors should follow GMP-style fill practices and minimize in-batch variation. Our ranking formula rewards consistency explicitly via the lot-consistency input.
What a pass looks like
- Automated vial-filling process, documented.
- Batches of 1000+ vials tested multiple times across the run with tight σ on purity.
- Public claim of which GMP standard (if any) the fill-and-finish operation follows.
What a fail looks like
- Hand-filled vials without process documentation.
- Purity variance >4 percentage points across vials from a single batch.
Product identification
Every container a researcher receives must be traceable to its batch and to the test results for that batch. Anonymous or unmarked vials break the chain of verifiable quality and are not acceptable.
What a pass looks like
- Batch / lot number on every vial label, not just the outer packaging.
- Batch-specific COA linked from the product page, not a generic COA.
- Recall capability — ability to reach every customer who received a given batch if a downstream issue emerges.
- Verification code on COAs or in product listing, checkable against a named analytical lab.
What a fail looks like
- Unlabeled vials with only a paper insert.
- One generic COA covering all batches indefinitely.
- No documented recall capability.
Sincerity and transparency
Doctored Certificates of Analysis, misleading marketing copy, and therapeutic claims on research-use products are all disqualifying. The bar is that a vendor’s public statements reasonably match what a buyer will verify when they test.
What a pass looks like
- Product pages frame every compound as research-use only. No dosing protocols.
- COAs are sourced from a named independent lab, not self-generated.
- Public communications about quality match what independent tests show.
What a fail looks like
- Edited or forged COAs (immediate disqualification and public correction log entry).
- Therapeutic claims on product pages, testimonials framed as treatment outcomes.
- Inconsistencies between vendor-stated purity and independent test results without acknowledgment.
Steady testing volume
Quality control is a flow, not a snapshot. Vendors that actually stand behind their product can sustain a steady stream of samples reaching independent labs over time. Our temporal-decay input in the ranking formula rewards recency.
What a pass looks like
- At least monthly batch-to-lab cadence when production is active.
- Annual testing commitment disclosed publicly (e.g., "≥12 per peptide per year").
- Willingness to participate in our free customer-submission pipe and/or the Golden Sample program.
What a fail looks like
- A single COA from 2023 still being cited in 2026.
- Testing only at launch, nothing after.
- Refusal to participate in any independent testing flow.
What sits on top of these standards
- The composite ranking formula converts test data into a 0–100 score. Standards tell a vendor what we're measuring; the formula tells them how the measurements add up.
- The objection process is the formal route if you believe we've scored something wrongly.
- The corrections log is where every change we make to published data lives, alongside the original.
- The vendor intake is how to apply to be listed if you meet the standards above.