A discovery system in public
Knowledge should show how it earned its status.
Synaplex is a working research system: it turns external sources into typed conjectures, registers falsifiable Claims before observation, and publishes findings only when Evidence and an accepting Decision exist.
0
Findings
No claim has yet completed the full Evidence-to-Decision standard. That is a result of the publication threshold, not missing page content.
4
Research records
Active, blocked, withdrawn, and invalidated work remains inspectable under stable identities.
18
Typed sources
Primary material from seven disciplines, retained with retrieval time, digest, source class, and access terms.
How the system moves
- Sourceinput, never Evidence
- Conjectureanalogy plus disanalogy
- Claimregistered before observation
- Evidencetyped and claim-specific
- Decisionaccepts, continues, or disposes
Current movement
Research does not disappear when it stalls.
A blocked study remains distinct from an invalidated one. A withdrawal is not a negative finding. Inline labels preserve those differences.
What has been learned
Engineering lessons, kept out of the findings lane.
Operational outcomes are useful, but they are not scientific Evidence. Each case names its limits and links to its source artifacts.
The build refuses unsupported findings
Canon validation and publication guards run before every site build, rejecting results without a valid Decision-to-Evidence chain.
Read the case →engineering case studyOne digest, two public readers
A deterministic projection is generated once, consumed by human pages and JSON clients, and deployed as an immutable Pages release.
Read the case →