SciencePaper 3
Protocol Paper
The current manuscript boundary for the Triple-Falsifiability Publication Protocol.
Implementation-aligned truth
The runtime now supports verifier-bound publication, sentence-level provenance, honest absence, and multi-axis scientific disclosure. The documentation and protocol framing now reflect that current product truth rather than an older speculative narrative.
Proof boundary
The protocol paper is stronger than the runtime docs, but it is still bounded. T.1 is empirically closed. T.4 now has archived second-domain verifier proof under the current protocol definition. T.2 is not closed. T.3 has strong repo-side evidence and multi-axis probe coverage, but the final synthesis still depends on explicit user acceptance rather than further repo density alone.
Current threshold map
The science section now includes an explicit T.1–T.4 status page plus stronger internal drafts for T.2, T.3, and T.4. Those materials make the remaining gap narrower, but they do not convert user closure authority into repo authority.
ECAM-X
runtime-closed in 1.0
Publication-required attribution validity. Claims must bind to evidence and survive verifier scrutiny.
AEI-Delta
stack and disclosure path active
Explanation stability when measured. The system discloses honest absence rather than inventing stability where none was measured.
TCA
stack and disclosure path active
Human reliance calibration as a story-scoped, audit-log-based disclosure rather than a convenience score.
What the paper can already say honestly
- The publication path is artifact-first and verifier-compatible.
- Missing scientific measurements stay explicitly disclosed as not measured.
- No story-level super-score overrides the underlying axes.
- Adversarial probe infrastructure now exists for all three axes.
- The current runtime/product path is stronger than a prototype claim.
What the paper must still not overclaim
- No claim that one second-domain proof automatically implies universal market-generality across arbitrary domains.
- No claim that T.2 is closed before the user pronounces the formal draft sufficient.
- No claim that repo-side T.3 runbooks alone equal final scientific closure.