Falsiflow helps make high-risk technical claims harder to advance without evidence. It does not create experimental facts. It audits the project config, evidence rows, source-file provenance, derived metrics, bundle manifests, and acceptance rules supplied by the user.
Falsiflow can report that a configured claim is ready under the evidence gates you defined. It can also show why a claim is blocked, which files support the claim, and whether a portable evidence bundle verifies against its manifest.
Falsiflow output is not proof of:
- biological safety
- clinical safety or efficacy
- regulatory compliance
- material biocompatibility
- experimental truth
- commercial readiness
- absence of hidden confounders or measurement error
Use Falsiflow as an audit and handoff layer, not as a substitute for independent experimental validation, expert review, legal review, clinical review, or regulatory review.
- Keep raw source files, chain-of-custody records, and metadata accurate.
- Define acceptance rules before interpreting results where possible.
- Preserve failed and blocked evidence instead of deleting inconvenient rows.
- Verify received bundles before relying on them.
- Document assumptions, controls, measurement limits, and known missing data.
- Avoid presenting
claim_readyas a universal truth outside the configured project scope.
For biomaterials, wetware, neural interfaces, medical devices, cell assays, or other safety-sensitive work, Falsiflow should only support planning, traceability, and review. Decisions that affect people, animals, patients, regulated products, or paid customer claims require independent expert judgment and appropriate external validation.