Verifying Self-Created Objectives — Reproducibility Container, Protocol Source, and Pre-Registration Archive (eve-repro / eve-protocol)

Versioned, citable reproducibility archive accompanying the three-paper release Verifying Self-Created Objectives by Matija Ludvig (Qualion Intelligence, a research programme of AWELON d.o.o., Slovenia): Paper I (the proposer-agnostic C1–C6 falsification protocol, study ladder, and measured adversarial false-accept rate), Paper II (deployment: a pre-registered gate on the production substrate, blind cross-validated external forecasts, and an externally-confirmed discovery), and the standalone letter A Measured False-Accept Rate for a Proposer-Agnostic Falsification Battery. Contents. • eve-repro — a frozen, one-command containerized image (pinned seeds, source and output hashes) that regenerates the pass pattern of every pre-registered ("starred") study, including the adversarial false-accept measurement (7.2×10⁻⁴ → ≤5.9×10⁻⁵), the constructed-bank operating characteristic, and the deployed Study 5 gate. • eve-protocol — released source for the C1–C6 detectors, the scaffold-identification and cross-controller harness, HomeoGrid, EveWorld, the removal-mode diffs, and the production components the deployed mechanism traverses. • Every SHA-256 hash-locked pre-registration and relationship-and-interest disclosure document, the OpenTimestamps (.ots) proofs anchoring them to the Bitcoin blockchain before unsealing, and MANIFEST.sha256 (the machine-generated digest list; its own digest is anchored).