VSO releases a reproduction kit for experimental VAP research
VeritasChain Standards Organization (VSO) has published a reproduction kit for its research on evidence submission and an OS-isolated mock execution gate. It gives external evaluators the environment requirements, execution steps, expected results and reporting format needed to run the fixed implementation themselves.
Research release · Experimental extension remains unadopted
Independent third-party reproduction has not been established. The results below come from internal testing and CI; they do not establish formal adoption, production readiness or full VAP conformance.
From a published implementation to a repeatable evaluation
The research separates cryptographic integrity from whether declared evidence requirements are satisfied. The kit connects the existing verifier, pre-record gate and OS-isolation experiment without changing their implementations. After environment setup, one command runs the existing regression and scenario suites, generates fresh synthetic evidence and rechecks signatures, hashes and bindings using the frozen code.
What internal validation confirmed
| Check | Recorded result |
|---|---|
| Package regression tests | 27 passed |
| Existing regression tests | 138 passed: 28 OS + 46 gate + 64 verifier |
| Total regression tests | 165 passed |
| Synthetic scenarios | 16 OS cases, 15 gate scenarios and 14 verifier cases matched their expected results; verifier cases were evaluated in both modes. |
| Saved evidence | 11 gate evidence bundles and the OS evidence were reverified. |
CI built and extracted the ZIP, checked its checksums, executed it against a separate checkout of the frozen subject and rechecked the saved outputs. “Matched” means the observed result agreed with the stated expectation. Negative controls intentionally demonstrate successful bypass when a route and an unauthenticated mock API are provided; these results are not a blanket claim that bypass is prevented.
The kit and the evaluated implementation are separate revisions
| Purpose | Commit |
|---|---|
| Reproduction subject: verifier, gate and OS experiment | c8ccf97 |
| Merged reproduction kit | b1af0dc |
| Final kit HEAD checked by CI | 69142d9 |
The package leaves current VAP v1.2, DOC-SYNC, K3, the existing verifier, existing gate and OS-isolation experiment unchanged. The verifier’s current-specification and experimental modes remain separate. Merging the kit does not assign a formal proposal number or adopt the extension.
Run it in an evaluator-controlled Linux environment
- Prepare a fresh, disposable native Linux host or Linux VM, then follow the setup and execution instructions. Git, Python 3.12+, rootful Docker Engine and Compose v2 are required.
- Obtain the fixed kit and the separately pinned subject, verify the source and run the preflight checks before executing the suites. The reference run used Ubuntu 24.04 x86_64, Python 3.12.15, Docker 28.0.4 and Compose 2.38.2; deviations are recorded.
- Retain the complete output and complete the evaluator report, including failures, incomplete steps and environment differences. Rechecking saved evidence alone is not a new OS execution.
The setup trusts the operator, host, Docker daemon and controller. It uses synthetic data, mock services and test keys; mock receipts do not establish independent time attestation. Initial setup needs external package and container registries. The next milestone is a separate evaluator’s actual execution and report, not an additional internal success count.